_..._ Natural Environment Research Council

-2 Institute of Geological Sciences

-

--

Mineral Reconnaissance Programme Report

c-

-

_.a

-

A report prepared for the Department of Industry

--

This report relates to work carried out by the British Geological Survey.on behalf of the Department of Trade I-- and Industry. The information contained herein must not be published without reference to the Director, British Geological Survey. I- 0. Ostle Programme Manager British Geological Survey Keyworth ._ Nottingham NG12 5GG I

No. 72 I A geochemical drainage survey of the Preseli Hills, south-west Dyfed, Wales I D I_ I

BRITISH GEOLOGICAL SURVEY

Natural Environment Research Council

I Mineral Reconnaissance Programme Report No. 72 A geochemical drainage survey of the I Preseli Hills, south-west Dyfed, Wales

Geochemistry I D. G. Cameron, BSc I D. C. Cooper, BSc, PhD Geology I P. M. Allen, BSc, PhD Mneralog y I H. W. Haslam, MA, PhD, MIMM

$5 NERC copyright 1984 I London 1984 A report prepared for the Department of Trade and Industry

Mineral Reconnaissance Programme Reports 58 Investigation of small intrusions in southern Scotland

31 Geophysical investigations in the 59 Stratabound arsenic and vein antimony Closehouse-Lunedale area mineralisation in Silurian greywackes at Glendinning, south Scotland 32 Investigations at Polyphant, near Launceston, Cornwall 60 Mineral investigations at Carrock Fell, Cumbria. Part 2 -Geochemical investigations 33 Mineral investigations at Carrock Fell, Cumbria. Part 1 -Geophysical survey 61 Mineral reconnaissance at the Highland Boundary with special reference to the Loch Lomond and 34 Results of a gravity survey of the south-west Aberfoyle areas margin of Dartmoor, Devon 62 Mineral reconnaissance in the Northumberland 35 Geophysical investigation of chromite-bearing Trough ultrabasic rocks in the Baltasound-Hagdale area, Unst, Shetland Islands 63 Exploration for volcanogenic sulphide mineralisation at Benglog, North Wales 36 An appraisal of the VLF ground resistivity technique as an aid to mineral exploration 64 A mineral reconnaissance of the Dent-lngleton area of the Askrigg Block, northern England 37 Compilation of stratabound mineralisation in the Scottish Caledonides 65 Geophysical investigations in Swaledale, North Yorkshire 38 Geophysical evidence for a concealed eastern extension of the Tanygrisiau microgranite and its 66 Mineral reconnaissance surveys in the Craven Basin possible relationship, to mineralisation 67 Baryte and copper mineralisation in the 39 Copper-bearing intrusive rocks at Cairngarroch Bay, Renfrewshire Hills, central Scotland south-west Scotland 68 Polymetallic mineralisation in Carboniferous rocks 40 Stratabound barium-zinc mineralisation in Dalradian at Hilderston, near Bathgate, central Scotland schist near Aberfeldy, Scotland; Final report 69 Base metal mineralisation associated with 41 Metalliferous mineralisation near Lutton, Ivybridge, Ordovician shales in south-west Scotland Devon 70 Regional geochemical and geophysical surveys in 42 Mineral exploration in the area around Culvennan the Berwyn Dome and adjacent areas, North Wales Fell, Kirkcowan, south-western Scotland 71 A regional geochemical soil investigation of the 43 Disseminated copper-molybdenum mineralisation Carboniferous Limestone areas south of Kendal near Ballachulish, Highland Region (south Cumbria and north Lancashire) 44 Reconnaissance geochemical maps of parts of 72 A geochemical drainage survey of the Preseli Hills, south Devon and Cornwall south-west Dyfed, Wales 45 Mineral investigations near Bodmin, Cornwall. Part 2-New uranium, tin and copper occurrence in the Tremayne area of St Columb Major 46 Gold mineralisation at the southern margin of the Loch Doon granitoid complex, south-west Scotland 47 An airborne geophysical survey of the Whin Sill between Haltwhistle and Scats’ Gap, south Northumberland On 1 January 1984 the Institute of Geological Sciences 48 Mineral investigations near Bodmin, Cornwall. was renamed the British Geological Survey. It continues Part 3-The Mulberry and Wheal Prosper area to carry out the geological survey of Great Britain and Northern Ireland (the latter as an agency service for the 49 Seismic and gravity surveys over the concealed government of Northern Ireland), and of the surrounding granite ridge at Bosworgy, Cornwall continental shelf, as well as its basic research projects: it 50 Geochemical drainage survey of central Argyll, also undertakes programmes of British technical aid in Scotland geology in developing countries as arranged by the 51 A reconnaissance geochemical survey of Anglesey Overseas Development Administration.

52 Miscellaneous investigations on mineralisation in The British Geological Survey is a component body of the sedimentary rocks Natural Environment Research Council. 53 Investigation of polymetallic mineralisation in Lower Devonian volcanics near Alva, central Scotland

54 Copper mineralisation near Middleton Tyas, North Bibliographic reference Yorkshire Cameron, 0. G., Cooper, C. C., Allen, P. M., and 55 Mineral exploration in the area of the Fore Burn Haslam, H. W. 1984. A geochemical drainage survey igneous complex, south-western Scotland of the Preseli Hills, south-west Dyfed, Wales. Mineral Reconnaissance Programme Rep. Br. Geol. Sun/. , No. 7 2 56 Geophysical and geochemical investigations over the Long Rake, Haddon Fields, Derbyshire 57 Mineral exploration in the Ravenstonedale area, Printed for the British Geological Survey by Four Point Cumbria Printing

I I I

I CONTEN‘IS SUMMARY 1 INTRODUCTION 1 1 Previous Work 1 GEOLOGY AND MINERALISATION 2 Stratigraphy 2 Cambrian 2 8 Ordovician: Arenig-Llanvirn 2 Ordovician: Llandeilo-Ashgill 4 Environment of deposition 4 Intrusions 5 Structure I Glaciation s Mineralisation 5 8 SAMPLING AND ANALYSIS 5 INTERPRETATION OF RESULTS 5 Frequency distribution 5 Definition of anomalies 6 Mineralogical examination of panned concentrates 7 8 Major sources of element variation 7 . 8 Assessment of anomalous results 11 CONCLUSIONS AND RECOYYENDATlONS 23 ACKNOWLEDGEMENTS 23 8 REFRRENCES 23 APPENDIX 1 Analytical results (pocket) 8 APPENDIX 2 Greyscale regional variation maps (pocket) FIGURES 1 Location of the survey area 2 1 2 Simplified geological map of the Preseli Hills 3 3 Diagrammatic summary of most highly significant inter-element associations 7 I 4 Copper in stream sediment 8 5 Copper in panned concentrate 9 6 Lead in stream sediment 14 7 Lead in panned concentrate 15 8 8 Zinc in stream sediment 17 9 Zinc in panned concentrate 18 10 Barium in stream sediment 19 11 Tin in panned concentrate 20 12 Boron in stream sediment 21 I 13 Sub area location map 22 1 Ordovician stratigraphy 4 Summary of anal&&al-results 6 Threshold levels and class intervals for anomalous results 10 8 Summary of highly significant inter-element correlations 12 8 1 .I SUMMARY INTRODUCTION A geochemical drainage survey at a density of 1 sample A reconnaissance drainage survey of the Preseli Hills in per km2 was carried out across the Preseli Hills, south- Dyfed, South Wales, was carried out as a means of west Dyfed. Stream sediment, water and panned con- assessing the mineral potential of the area and to centrate samples were collected from each of 358 sites, provide baseline geochemical data on the region. The and Cu, Pb, Zn, Ba, Fe, Mn, Co, Ni, V, Cr, B, Zr, As, MO work formed part of a larger project to assess the and Sn were determined in sediment, Cu, Pb, Zn, Ba, Fe, mineral potential of Precambrian and Lower Palaeozoic Mn, Ti, Ni, Sn, As, Ca, Ce, Sr, Sb, Zr, U and MO in rocks of southwest Dyfed (Pembrokeshire). The geo- panned concentrate and Cu and Zn in water. chemical drainage survey was restricted to this area, as From a study of regional variation patterns and previous studies (Allen and others, in prep) showed that multivariate statistical analysis, the main sources of this technique was ineffective in lowland areas, failing geochemical variation in the data were found to be to detect a substantial area of copper enrichment associ- bedrock lithology, mineralisation, contamination and ated with disseminated porphyry style mineralisation. hydromorphic processes. Strong geochemical signatures The survey area (Figure 1) consists of the east-west were shown by dolerite intrusions and by acid volcanics ridge of the Preseli Hills rising to over 500 m, the of the Fishguard Volcanic Group. Dark mudstones of the surrounding lower plateau of 100 to 200 m in height and D. murchisoni Beds and Sealyham Volcanic Series also a small section of coastal platform 50 to 60 m in height. showed characteristic geochemical features whose Most of the area was subject to glacial and periglacial impact was limited by their restricted and sinuous out- processes during the Pleistocene era. The plateau area is crops. A feature of most of the area is the presence of gently rounded in the east, with several fairly broad monazite nodules, generating high levels of rare earth river valleys running north-south, while the west is more elements and uranium in the panned concentrates. These severely incised where sub-glacial channels have cut are particularly high over the Llandeilo-Ashgill sedi- deep narrow gorges in p&es. The coastal platform has mentary rocks. also been cut into by ice-marginal or sub-glacial streams Geological thresholds were established using cumula- forming several deep gorges (Jones, 1965). Most of the tive frequency plots and percentile division. Anomalies lower plateau and coastal platform is farmed, while the due to contamination could be discriminated satisfactor- higher ground is mainly rough grass and heath. High ily from those arising from mineralisation only by field rainfall assists peat formation, and large areas of ill- observation and mineralogical examination of panned drained wetland are present. Rock outcrop is poor concentrates. In major streams crossing lowland areas because of extensive till and head deposits (Bradley, and in the vicinity of roads, widespread contamination 1976) being largely restricted to the sides of incised proved to be the source of most large base metal river valleys and to the characteristic cams of the main anomalies. ridge. Anomalies related to known mineralisation are located in the southeast of the area, around Llanfyrnach in the Taf valley. Anomalies reflecting hitherto unre- Previous work corded bedrock mineralisation may be present in the Geologically, the area covered by the drainage survey following areas: has been studied as a whole only by W. D. Evans (1938, 1940, 1945), most other mapping having concentrated on a1 Llanfyrnach; anomalies for Cu, Pb, Zn and Ba are the igneous rocks to the west of the Preseli Hills. Part of more extensive than can be accounted for by the the area was mapped by Cowper Reed (18851, followed known vein mineralisation. High levels of other ele- by Part (19221, who described the igneous rocks of the ments suggest an association with shales. Preseli Hills. The area east of Fishguard has been remapped by Lowman (1977; Lowman and Bloxam, 19811, b) Crosswell-Crymmych; barium and base metal anoma- while further west Elsden (19051, Thomas and Cox (19241, lies are associated with the Fishguard Volcanic Group Cox (1930), Williams (19331, Thomas and Thomas (1956) and overlying pyritiferous dark mudstones of the D. and Bevins and Roach (1979a, 1979b) studied primarily murchisoni Beds. The geological setting of these the igneous series of rocks between St David’s and rocks, deposited in a sulphurous environment on the Fishguard. 0. T. Jones (1912) covered the Cardiganshire flanks of an acid volcanic pile, suggests that massive area and D. C. Evans (1906) Western Carmarthenshire. sulphide deposits are a possible target here. Locally, The Quaternary geology of the St. David’s peninsula and weak arsenic anomalies, old trials and ancient mining the Preseli Hills has been studied by Charlesworth records suggest that there is also some potential for (1929), Griffiths (1939), 0. T. Jones (1965) and John Au mineralisation associated with pyrite in veins and (1965, 1967, 1970, 1971). mudstones. Some geochemical studies have been made of the area. The Wolfson Geochemical Atlas, (Imperial College cl Minas Dinas and Pentre Ifan; small amounts of gold of Science and Technology, 1978) can be used to deline- and chalcopyrite were found in panned concentrates. ate broad regional trends and Urquiddi-Barrau (1973) used this as baseline data for further work in Wales, d) Very small amounts of cassiterite were found in including soil traverses across four areas in the Preseli panned concentrates from several localities, but no Hills. No anomalies adjudged to have any economic source was readily identifiable. potential were discovered. Bradley and others (1978) also The anomalies in the Llanfyrnach and Crosswell areas covered part of the Preseli area in a survey of the were considered to. merit more de tailed investigation and distribution of trace elements in northwest are the subject of further work by BGS. Pembrokeshire soils. 1

Figure 1 Location of the survey area

Regional gravity and aeromagnetic surveys have been the Ordovician succession Evans (1945) recorded the carried out by the (British Geological Survey) (Cornwell graptolite Dfdymograptus eJttensus from the upper part. and Cave, in preperation). Sealyham Beds GEOLOGY AND MIBJBRAidSA’IION Most of the southern half of the area is underiain by The rocks, predominantly Ordovician in age with small unfossiliferous blue-black slate and cleaved mudstone, faulted inliers of Cambrian sedimentary rocks in the which Thomas and Cox (1924) considered were equivalent west, make up a sedimentary and volcanic succesion to the upper part of the Tetragmptw Shales. They are some 1.3 km thick, with some contemporaneous intru- most likely to be uppermost Arenig in age. In other parts sions (Figure 2). The rocks were folded and, according to of the county, the Tetmgmptus Shales mark a quiescent Bevins (19781, metamorphosed to prehnite-pumpellyite period between the early Arenig volcanism, represented facies during the Caledonian orogeny. The area lies close by the Treffgarne Andesites and the prolific Llanvirn to the Hercynian and the rocks are almost certainly volcanism. affected by earth movements related to this orogeny. Much of the area is thought to have been glaciated Sealyham Volcanic Group during the Pleistocene. Several small, generally fault-bounded, areas of volcanic rocks in the southern Preseli Hills were correlated by Evans (1945) with the Sealyham Volcanic Group (Series, The stratigraphical succession for Ordovician rocks given in the original publications). The group is no more than by Evans (1945) for the eastern half of the area and 80 m thick and consists of rhyolites and minor tuffs. Part Lowman and Bloxam (1981) for the western part, is (1922) gave the first descriptions of these rocks, which shown in Table 1. include perlitic and brecciated soda-rhyolites and soda- trachytes. The volcanic rocks are generally believed to be upper Arenig or lowermost Llanvirn in age, conform- The oldest rocks in the area occur in small faulted inliers ably followed by mudstones containing the diagnostic in South Fishguard and along the Gwaun Valley. Low man lower Llanvirn graptolite Didymograptus bifidus. How- (1977) correlated them with the Lingula Flags. A larger ever, some doubt about this age is indicated by Part inlier north of Ysgubor Mountain (SM 965 315) consists of (1922) and the rocks may be younger, possibly equivalent the Llanfair Beds, which are of either Cambrian or to the Fishguard Volcanic Group. Arenig age. Didymograptus bifidus Beds ordovician: Ares@-- The greater part of the Preseli Hills is underlain by blue- Foe1 Tyrch Beds black mudstones and slates with thin bands of tuffaceous These rocks, not distinguished on Figure 2, are confined or tuffitic rocks in a succession likely to be over 300 m to the northeastern end of the Foe1 Tyrch anticline. thick. Very few fossils have been found in this succession They consist, in ascending order, of splintery ‘chinastone’ (which is generally attributed to the lower Llanvirn) and ashes, blue-grey and greenish grey slates, and ashy only at Cam bleini (SN 142 327) are they undoubtedly of mudstones with interbedded feldspathic sandstones. the bifidus zone (Evans, 1945). As a result of the doubt They resemble the lower Arenig Brunel Beds, which, about their age, Lowman and Bloxam (19811, preferred to elsewhere in Pembrokeshire, occur at or near the base of name the formation Gwaun Valley Slates.

2 ___ Fawr S&B Mvdnm Shah Hondra Sh&r Beda

ORDOVICIAN -

_____-_ Gdogaoalboundaw

s-w FauIta. thnmaand linamona intafpmtad aa faulra (no lndimlon of dobmmr0w glvan)

Gaologv comrulad fromW.D.Evam (1046). LOWIM & Bloom (1981) and field da0 comr~~ied bv W.D.Evana on &inch rnapm lodgad wth IGS

Figure 2 Simplified geological map of the Preseli Hills

Fishguard Volcanic Group The thickest development of this group is on Evans (1945) noted that the D. bifidus Beds thinned Strumble Head, where about 1.8 km of acid and basic eastwards, more than could be accounted for structurally volcanic rocks were divided into three major divisions by and suggested that the Fishguard Volcanic Group, which Cox (1930) and Thomas and Thomas (1956). East of follows the D. bifidus Beds, may overlie them, Fishguard, Lowman and Bloxam (1981) were unable to unconformably, here. On the Pen Caer peninsula, how- subdivide the Fishguard Volcanic Group into lithostrati- ever, Cox (1930) and Thomas and Thomas (1956) did not graphic units. The group thins towards the east, being no question the conformable nature of the junction. In more than 870 m thick near Newport (Lowman, 1977) and addition, on the strength of specimens of D. bifidus in 90 m in the Preseli Hills (Evans, 1945). the shales overlying the volcanic rocks, on the east of Within the group, Lowman and Bloxam (1981) recog- Fishguard Harbour, Thomas and Thomas (1956) referred nised vitric pyroclastic flows, pyroclastic fall deposits, to the volcanic rocks as lower Llanvirn in age. Lowman epiclastic deposits, rhyolitic lavas, basaltic pillow lavas (19771, however, disputes this, having identified and hyaloclastites. The tuffs show evidence for both D. murchisoni Zone fossils in mudstone within the subaqueous and subaerial eruption and deposition. Acid volcanic group on Castle Point. rocks predominate, the only basic extrusives forming a

3 Table 1 Ordovician stratrigraphy of the Preseli Hills

Eastern Area Western Area Other Areas

Freni Fawr Bed Ashgill Glogue Slates to Mydrim Shales Llandeilo Hendre Shales Hendre Shales Castle Point Beds

D. murchisoni

Llanvirn Fishguard Fishguard Volcanic Volcanic Series Group

D. bifidus Beds Gwaun Valiey Slates

Sealyham Volcanic Series

Arenig Sealyham Beds Tetragmptus shales

Foef Tyrch Beds Brunel Beds

small outcrop near Carn Gelli (SM 981 376). Alloclastic to Evans (19451, from over 50 to 100 m thick and are dykes occur within welded tuff locally. Evans (1945) overlain unconformably by the arenaceous Glogue Slates. describes ‘ashy’ mudstones within the group in the The Freni Fawr Beds consist of interbedded turbiditic Preseli Hills. sandstones and mudstones. A conglomerate consisting of Chemical studies carried out west of the area, indi- sandstone, mudstone and shale pebbles in a silicified cate thoieiitic affinities for the basic intrusions and matrix is recorded in the succession at the type locality. iavas. Bevins and Roach (1979a) argue, using only the trace-element chemistry,_ that the rhyolites on Strumble Head are also tholeiitic. Lowman and Bloxam (19811, on Envinwunent of deposition the other hand, claim that the acid rocks are cab Though little is known about the earliest Ordovician alkaline. rocks in this area, the equivalents of the Foe1 Tyreh and Brunei Beds elsewhere suggest deposition in shallow, turbulent, littoral or sub-littoral waters. Much of the Didymograptus murchisoni Beds elastic material in them is of volcanic origin, derived In the Preseli Hills, Evans (1945) described 15 to 25 m of from either the Treffgarne Andesites or earlier volcanic dense black, commonly pyritous, mudstones and shales piles. with upper Llanvirn D. murchisoni Zone fossils, overlying Deeper water sedimentation began during the Arenig, the Fishguard Volcanic Group. The uppermost ‘ashy’ from which time the predominant lithology was mud- mudstones in the volcanic group pass either conformably stone. A profound change in basinal development took or with a minor non-sequence into the black mudstones. place in late Ordovician times, with the onset of turbid- The D.murchisoni Beds thin westwards and are over- ite sedimentation, represented by the Freni Fawr Beds. stepped by. the Hendre Shales of Llfrndeilo age, which Until then, several minor unconformities suggest that rest directly on the volcanic group near Dyffryn-benglog the volcanic rocks represented by the Sealyham and farm (approx. SN 107 385). Near Fishguard, Lowman and Fishguard volcanic groups were erupted from a series of Bloxam (1981) claim that the Fishguard Volcanic Group volcanic islands. Evidence of subaerial erosion and itself is upper Llanvirn in age. deposition is present in the Fishguard Volcanic Group. The age of the Sealyham Volcanic Group , commonly put at Arenig-Llanvirn, is not certain. The presence of Ovdovicim Lhindeilo-m D. murchisoni Zone fossils, within and above the Calcareous flagstones, the Castle Point Beds, considered Fishguard Volcanic Group, indicates that the latest to be Llandeilo in age by Cox (1930) and Thomas and volcanism in the area is of this age. The Fishguard Thomas (1956), unconformably overlie the Fishguard Volcanic Group is thickest on Strumble Head, probably Volcanic Group immediately east of Fishguard Harbour. near the centre of the eruption and thins eastwards so They are followed by the Hendre Shales, some 60 m that the outcrop in the Preseli Hills is probably part of thick, which Evans (1945) suspected lie unconformably on the lower flanks of the volcano. There is evidence that, both the volcanic group and the D. murchisoni Beds. west of Newport, the uppermost volcanic rocks have They are monotonous, brown-weathering, blue-grey, been above sea level. To the southeast of Newport, the slightly calcareous, poorly fossiliferous mudstones and volcanic rocks are most likely to have been erupted or shales. deposited below sea level and there are sedimentary Rocks of Caradoc and younger age are represented by intercalations throughout the volcanic group in this area. the Mydrim Shales, Glogue Shales and Freni Fawr Beds in Formation of the black, pyritous mudstones of the the northeastern and eastern parts of the area. The D. murchisoni Beds may have been enhanced by Mydrim Shales consist of pale-weathering blue-black volcanogenic sulphide emission. They provide the most shales and mudstones, locally fissile and crowded with promising environment for the formation of exhalative poorly preserved graptoiites. The shales are, according sulphide deposits.

4 r Intrusion Gwyn (SN 172 365), tried unsuccessfully for lead ore in Lowman (1977) described intrusions of microgranite near the late nineteenth century, and Fron Las (SN 166 340), Newport, intruding dolerite associated with the an unsuccessful trial for copper. At Fron Lwyd Fishguard Volcanic Group and he notes some microdior- (SN 178 339) a small mine was operated from 1864 until ite intrusions emplaced in both the volcanic group and 1875, exploiting east-west striking veins containing lead the underlying rocks. Most numerous, however, are and copper ores. Production here was limited and no dolerite intrusions, mostly forming sills, confined traces remain of any of these three workings. entirely to the lower part of the Fishguard Volcanic Evans (1938, 1940, 1945) reports sulphide mineralisa- Group and formations below it. Evans (1945) describes tion at various localities in the black shales and mud- the dolerites in detail. They are most likely to be stones of the D. bifidus Shales, D. murchisoni Shales and products of the Fishguard Volcanic Group magmatism, Mydrim Shales as weIl as tuffs of the Fishguard Volcanic though Evans (1945) considered them to post-date the Group. Some localities include pyritised D. bifidus main folding. Shales on Carn Goedig and Carn Breseb; tuffs with pyrite in the valley of the Afon Wutehook, in addition to pyritised D. murchisoni Shaleskthe same locality. This strlJcture latter Iithology is noted as highly pyritised elsewhere The structure of the area has been interpreted by Evans (see above). (1945). He described a series of major, broad, east- trending folds called, from north to south, the Nevern SAMPLING AND ANALYSIS Anticline, Brynberian Syncline, and Crug-yr-Hwch and Sampling was carried out at a nominal density of one Foe1 Tyrch anticlines. Minor, in places tight, folds with the same trend are common and the southern half of the sample site per km*, but the relative lack of surface drainage in the west of the area reduced the density to area has been further complicated by a series of south or as low as 0.7 site per km*. At each site, a sediment, southeasterly directed thrusts. Rocks associated with the panned concentrate and water sample were collected. thrusts are either intensely sheared or tightly folded. Sediment samples were collected by wet sieving at site, through 8 mesh and 100 mesh nylon into a pan. The Glaciation Pembrokeshire was subjected to glaciation during resulting -100 mesh material was allowed to settle during collection of the other samples. Excess water was Pleistocene time, but the effects are not particularly decanted and the sample bagged in a ‘kraft’ paper bag obvious in most areas. The earliest glaciation of the era which was then air dried. In the laboratory, samples were resulted in Irish Sea ice covering the peninsula, giving disaggregated and ground for 30 minutes in a P5 ball rise to the Older Drift (Charlesworth, 1929). The second grinder. Cu, Pb and Zn were determined by Atomic and last glaciation followed a prolonged periglacial Absorption Spectrophotometry (AAS), after digestion of period but left relatively little evidence in the HiIls a 0.5 g sub-sample in hot concentrated nitric acid for proper; the area being in a marginal position to the main one hour. Other elements were determined by Optical ice sheet. The South Wales End Moraine, thought to Emmission Spectrography (OES), with the exception of mark the limit of the ice, was mapped along the coastal As for which a further subsample was prepared for platform by Charlesworth (1929). Local ice-related analysis by X-Ray Fluorescence Spectrometry (XRF), as features are found on the higher ground, and some perfglacial deposits are found in the valleys (John, 1972). outlined below. Panned concentrates were obtained by With the wasting of the ice, large amounts of outwash panning approximately 4 kg of -8 mesh stream sediment. They were examined on site for any evidence of mineral- material were deposited along the northern coastal isation. In the laboratory, the samples were dried and a plateau, and large valleys were cut by subgiacial and ice- 12 g subsample was taken and ground for 6 minutes with marginal channels. In the west, the large Gwaun- 3 g ‘Elvacite’ binder in a Tema miIl prior to pelletising Jordanston system was originally thought to have been and analysis by XRF for a range of elements. cut by overflow channels (Charlesworth, 1929), but is now thought to have been formed below the ice, Ni and Pb in panned concentrate results are enhanced by REE interference when Ce levels are high (>c.3000 indicating that the margin was further south than the ppm). Ba in panned concentrate results are depressed by ‘South Wales End Moraine’ (John, 1972). high levels of Ce. Water samples were collected in 30 ml polyethylene Mineralisation bottles, acidified in the field with 0.3 ml HC104 to Known workings and trials are restricted to the eastern prevent sorption of metals by the container walls and part of the survey area, some seven sites being known. subsequently analysed by AAS for Cu and Zn. . -The second most productive metal mine in South Wales was situated at Llanfyrnach in the valley of the Taf INTERPRETATION OF RESULTS (SN 225 316). It had a recorded production of 15 653 tons All analytical results are given in Appendix 1 and a of lead concentrate, 763 tons of zinc concentrate and summary is shown in Table 2. Results less than the profitable amounts of silver, during its operations detection limit (Table 2) were reported as half that between the middle of the eighteenth century and 1890 level, and these values were used for statistical analysis. (H-“3h-9 %971x Se veral veins were worked, all of which cut It proved impossible to subdivide sample populations on the black Mydrim Shales; no other rock type being the basis of background geology because of the rapid present on the dumps (Evans, 1940). Foster-Smith (1981) alternation of Iithotypes within catchments; so for initial states that the gangue minerals were quartz and calcite statistical interpretation the data were treated as a with galena, sphalerite and pyrite. Evans (1940) also single population. reports the presence of chalcopyrite. In close proximity For brevity, ‘in sediment’, ‘in panned concentrate’ and to the main Llanfyrnach workings are two trials at ‘in water’ are abbreviated to subscripts s, p and w LIwyn-yr-Hwrdd (SN 225 324) and Llwyncelyn respectively. (SN 232 314) which were made to find extensions of the Llanfyrnach veins. These were not successful. At Pont- y-Gafel, on a tributary of the Taf, a level was driven Frequencey distribution into a quartz vein cutting D. bifidus Beds. Little is Cumulative frequency probability graphs and histograms known of the object of the working, which was were plotted to examine element distributions abandoned. (Lepeltier, 1969; Parslow, 1974; Sinclair, 1976). Five types were identified: On the northern slopes of the Hills, immediately west of Crymmych, two trials are reported at Llanfair-nant- a) Normal, shown by Fe,, Vs and B,.

5 b) Lognormal, the largest group, consisting of Cu, Zn, e) Complex. Pb shows a complex form which can be Co, Ni, As and Zr in sediment, and Ti, Sn, As, Ca and Ce interpreted in Perms of three overlapping lognormal in panned concentrate. Some of these distributions are populations. only very approximately lognor mal. Cu and Zn in The full distributions of Sn, Sb, U and MO in panned sediment could be interpreted as consisting of a lower concentrates, Cu and Zn in water and MO and Sn in normal population and an overlapping lognor ma1 stream sediment were indeterminate as a result of heavy population. Cep and probably Snp populations consist truncation (Y&I% of samples

Table 2 Summary of analytical results in ppm for 358 stream sediment, water and panned concentrate samples.

Stream Mean Median Standard Maximum Minimum Geometric Geo. mean Geo. mean Detectfon sediments deviation mean +geo. dew. +2 geo. dev. Zimit cu 21 20 9 60 S 19 30 46 Pb 61 40 260 4700 20 40 68 11s Zn 177 150 103 1300 30 158 251 398 Ba 676 532 594 5191) 197 57s 912 114s Fe 46911 45800 11456 91900 13200 45709 58884 75858 Mn 3088 1780 3328 25700 124 199s 5012 12589 co 3s 25 32.8 261

Water cu* (0.01 0.13

Panned concentrates cu 40 17 203 3740 <6 18 47 120 6 Pb 168 42 1178 21884 <13 51 151 447 13 Zn 189 146 531 991s 24 141 257 468 3 Ba 349 242 827 14200 <27 145 646 2884 27 Fe 74401 72420 29297 234530: 13690 69183 104713 158489 Mn 2629 1240 3258 18840 210 1514 4169 11482 6 Ti 29488 18000 34902 241000 2310 17783 147863 128825 Ni 39 39 17.5 103 <3 3s 60 105 5 Sn 61 12 173 2455 <9 18 72 295 9 As 6 5 6.4 66 (2 4 11 26 2 Ca 5910 3640 6640 44840 200 3311 10233 31623 Ce 5727 1231 11774 84500 . 37 1259 7762 47863 21 Sb*

All results less than detection limits are set to half that value. * Variables in which >50 per cent of values are below detection limit.

6 . Basic intrusions

V 1 Acid Volcanics Cap Mop

Tip Fe, - Precipitates . MnP Nip ----- me---- m--m- r c \ l Base M_etal_ Mineralisation

. 1 * M”s Fes I cu 1 I\ s - I *\ I I Contamination Mudstones 1 ‘1 1 Bas i and 1’ t ’ ‘\ L-d Tuffs

Granitic / ?Heavy Mineral / Nodules / / Group / / / / f /

Figure 3 Diagrammatic summary of most highly significant inter-lement associations from the background population ocurred on the cumula- where necessary, by XRF and X-ray diffraction (XRD) to tive frequency plot (Sinclair, 1976). The higher sample aid identification of natural and artificial phases respon- population (defined as anomalous) contains a proportion sible for metal anomalies. of the background population, particularly close to threshold, but few samples belonging to the upper Major sources of Element Variation anomalous population will escape identification. Inter-element relationships were investigated using a Threshold levels set on this criterion are munder- variety of multivariate statistical methods and spatial lined in Table 3. Sn in panned concentrates and sedi- distribution plots. ments were treated as special cases because of their role Gross variations of the data across the area were as indicators of contamination: all results above the studied using greyscale maps (Appendix 2). Each symbol detecRjon limits, which coincided approximately with the on these maps represents the mean value over nine cells, maximum background levels likely to be encountered in each cell being 0.15 km *. The classes represented by the common rocks, were taken as anomalous. All Ce in symbols were normally constructed on the basis of the panned concentrate results except the lowest 5%, were (0, 30, 50, 70 and 90 percentile levels. Bearing in mind regarded as anomalous: there is a clear inflexion point on their limitations, these maps fulfill the purpose of the logscale plot at this level and the upper population drawing attention to relatively large areas characterised values would be considered anomalous in other compar- by particular levels of a given element. able areas, such as Anglesey (Cooper and others, 1982). Elements whose distributions approximated to log- Above the threshold level, anomalies were divided normal form and others whose skewness was reduced by into classes based on the 90, 95, 97.5 and 99 percentile the process, were log-transformed prior to applying levels (Table 31, which were used in plotting anomaly parametric statistical methods. Following the removal of maps for certain elements (Figures 4-12). highly truncated and suspect variables and setting very high significance levels,sensible results were obtained from multivariate statistical analysis despite the Mineralogical examination of panned concentrates theoretically unsound database, so these were used with Fifty nine anomalous panned concentrates were selected caution in the overall interpretation. for detailed examination in an attempt to identify mineral phases responsible for high metal levels and thereby indicate the most likely causes of anomalies. After removal of the ferromagnetic fraction, the 3.3 The following major sources of element variation, in sink fraction was sieved and the 30-60 mesh subsample the data set were identified by a combination of statisti- separated into fractions according to magnetic suscepti- cal analyses, mineralogical examination of the panned bility. Each fraction was examined microscopically and concentrates and spatial distribution plots (Appendix 2)

7 Copper in Stream Sediment.

KEY 0l 43- 48 ppm 0. 349 Ppm

Figure 4 Copper in stream sediment Copper in Panned Concentrate.

. < 50 ppm

0 50-l 00 ppm .y

. 101-170 ppm 0 .

0. 171.354 ppm -.7 0. ,355 ppm - i - --

Figure 5 Copper in panned concentrate

9 n Table 3 Threshold levels and class intervals for anomalous results, in ppm.

Percentile level

<90% 90% 95% 97.5% 99%

Sediments cu 2s 31 36 43 49 Pb 51 71 111 625 Zn 150- z?l 341 371 481 %a 9&l 1400 3100 3500 Fe 67950 72850 76850 Mn 1gJ.g 6900 9500 13250 15400 co 118 140 Ni 90 110 V 175 194 203 Cr 102 150 169 181 192 B 106.5 114 120.5 Zr 810 950 AS 45 60 MO 4.75 7.15 Sn 5.6 13.5 21.5 28 41

Water

CU 0.015 0.025 Zn 0.035 0.085

PaMedconcentrates ’ cu 50 (91%) 101 171 355 Pb -90 (80%) 201 401 951 1201 Zn 250 (92.5%) 302 441 615 Ba 750 (95.5%) 1100 2400 Fe 130000 160000 Mn 6350 8990 13550 16650 Ti 130000 170000 Ni 67 79.5 90 Sn 9.5 139 255 435 620 As 20.5 30.5 Ca 24200 33100 Ce 70 (5%) 17000 28500 40000 64000 Sb 12.5 Sr 131 151 181 211 ::I Zr 331 (60%) 761 1061 1701 2201 U 45 65 MO 6.5 10.5

relatedto geological, topographical and agricultural also form part of this association. Elements involved in features. A summary diagram of inter-element associa- this process form a clear group in R-mode cluster and tions is shown in -Figure 3; this is based. ._--_ on the cluster factor analysis models of the data and show a negative and factor analysis, and__ -- correlation matrix. relationship to Zrs because of contrasting modes of occurrence and concentration in the primary and secondary environments. Hydrous oxide precipitates Some metals, soluble in the acid groundwater of the peaty upland areas, are much less soluble in the more Contamination oxygenated and less acidic conditions which prevail in The mineralogical examination of panned concentrates the surface streams and groundwater at lower altitudes. showed tin to be present both as cassiterite and as This results mainly in the formation of hydrous artifi&l phases. The amounts of cassiterite found were, manganese and iron oxide precipitates although other however,very small (one or two grains) and all large tin metals may also precipitate out or be scavenged by anomalies were caused by contaminants such as tin cans precipitates (Nowlan, 1976; Reedman, 1979). The process and solder. It was concluded therefore that, with is illustrated by the formation of Fe and Mn rich caution, the tin content of samples could be employed as coatings on pebbles in upland streams. Although mainly an indicator of contamination in this area. Lead was affecting sediment results, panned concentrates may be found also to be present both in contaminants (mostly affected where coatings are not removed during the yellow or brownish glass) and in the form of galena, both rubbing and sieving process. Known geochemical phases being present together in some samples. High behaviour, strong inter-element associations, a contrast antimony levels correlated very closely with high levels between sediment and concentrate sample results and a of lead in contaminated samples and as no natural correlation of high results with peaty uplands and their antimony source was identified, antimony could be margin? rather than geology, indicates that the elements regarded as an indicator of contamination in this area. affected in this area are Fe, Mn, Co, As, Zn and Ni in Wire containing copper and zinc was seen -in many sediment and that Mn, Ni and Fe in panned concentrate panned concentrates. Since tin, copper, lead and zinc

10 occurred in both natural and artifical phases, a clearly MudstoneS, shales and tuffs of Arenig and Llanvirn age defined group of elements indicating the presence of High levels of Crs, Vs, Fes, Ni,, Cos, Ass, Bs, Mos, Zns, contamination could not be defined by multivariate Tip and Cus are characteristic. They strongly reflect the statistical methods. Mineralogical examination was the amount of shales and mudstones in the catchments. only effective means of discriminating between metal Some local subdivisions are evident: high levels of Ba,, anomalies related to contaminants and those related to are a prominent feature of the D. murchisoni mudstones mineralisation. and high levels of Cr and V are associated with the Sealyham Volcanics. High Ba levels, accompanied locally by V and Cr, are a fea aure of the Foe1 Tyrch Base metal mineralisation Beds. Again, these lithologies provide a source of metals Multivariate statistical methods were also relatively subject to hydrous oxide precipitation processes. Poor inneffective in delineating any mineralisation in this exposure and limited outcrop reduce the impact of area. This is a result of the elements involved having lithologies such as the D. murchisoni Beds. other sources of variation which accounted for a larger proportion of that variance. In addition, BaP results, Fishguard Volcanic Group often useful indicators of mineral’ tion, were variably Zr values are high over most of the outcrop, whilst depressed by Ce interference, m z ing any meaningful spatial distribution plots (Appendix 2) of Ti , Nip and Bs associations with other elements. Consequently R-mode show clearly defined zones of low values, Feflecting the cluster and factor analysis models yield a grouping of predominance of acid volcanics in the succession. minerali.tfg$ymetals (Pb c~;?J$m~&io;;P~ *se), due to both Llandeilo-Ashgill succession The lack of a distinct factor identifiable with These rocks display high levels of B,, Cep, Up and mineralisation suggests that it is of minor significance in erratically Zrs and Mos. Low levels of Ti , Mnp and Fe terms of geochemical variation in this area. Site inspec- are also seen. There are high values For a range o! tion and mineralogical examination were the main tools elements along the outcrop of the Hendre and Mydrim used for distinguishing anomalies caused by mineralisa- shales of the Taf valley which may originate from these tion. Mineral examination identified chalcopyrite in the lithologies or from associated mineralisation. Elements panned concentrate from one site and gold in another. In affected are: Cu , Pbsp, Znsp, Basp, Bs, Nip, Assp, Mosp, addition, gold was noted during the on-site examination Cep, Srp, Zrp an a Up* of another panned concentrate.

Assessment of anomalous results Monarite The survey area is divided into ten sub-areas on the basis Exceptionally high Ce values were found in panned of overall geochemical similarity and catchment concentrates over much of the area, and mineralogical boundaries (Figure 13). Anomalous results and examination showed this to be due to the presence of geochemical features in each sub-area are briefly monazite nodules in the concentrates. Research in other discussed. parts of Wales suggests that there is no single source for the nodules, that they occur dispersed through a great thickness of Ordovician and Silurian sedimentary rocks Afon Taf, Afon Cafe1 deposited in the Welsh Basin and that they have a pre- The catchment of the Taf contains Llanfyrnach silver- metamorphic origin (Cooper, Basham and Smith, 1983). lead mine, which worked galena and sphalerite in a The source rocks in Central Wales show Ce levels gangue of calcite and quartz (Foster-Smith, 1981) Spatial indistinguishable from typical background levels in distribution maps of several elements (Cus, Pbs, Zns, similiar lithologies. Separation from these rocks showed Mo , Ass, Bs, Cup, Pbp, Zn , Ba , Nip, Mop, Asp, Snp, the presence of 400-600 nodules in each 15 kg sample, Sr z , indicate a broad zone o! ano pnalies which cannot be roughly equivalent to 18-27 ppm Ce in a whole rock so Pely accounted for by dispersion from the worked analysis (Read, in prep). The high levels in the mineralisation. In addition, Cep, Zrsp and Up values are concentrates are a product of upgrading by the present high over this area as well as the remainder of the weathering cycle and the sampling process. Detailed Llandeilo-Ashgill Series. The mine dumps are situated on work on the origin of the nodules will be reported the west bank of the Taf and contribute material to the elsewhere (Cooper, Basham and Smith, 1983; Read, in river, giving rise to Pb and Zn anomalies downstream. preparation). Apart from this obvious source, there are other Pb and Up anomalies are also related to high concentrations Zn anomalies on the Taf and its tributaries. Above the of monazite in panned concentrates. The presence of U mine workings some anomalies, e.g. at SN 2196 3250 and Th in nodules was confirmed by XRF analysis during (no. 184), may be partly due to observable mineralogical examination. The highly negative contamination, but galena was recorded in the panned correlation of Cep with BaP is the product of analytical concentrate at SN 2260 3229 (no. 129), suggesting the interference. presence of further base metal mineralisation. Base is also high and baryte is recorded in several panned Lithology concentrates, e.g. SN 2164 3043 (no. 1301, SN 2254 3128 Some of the rock groups outcropping in the area provide (no. 131) and, possibly, SN 2005 3318 (no. 116). In other distinct geochemical signatures and give rise to anomalous samples, Ba, is usually greater than Bap, anomalous levels of some elements. suggesting another source for the barium, such as feldspar or muscovite, but analytical interference from Ce suppressing Bap could also be the cause. For example, Dolerites intruded into the Lower Llanvirn at SN 2254 3128 (no. 131), baryte was found during These rocks give rise to high Tip values, associated with examination of the panned concentrate even though Bap Mnp, Nip, Fep, Ca and, locally, Mop, due to the was reported below the detection limit (Ce 54000 ppm). presence of abundan P Fe-Ti oxides and silicate phases Thus baryte may not be revealed as anomalous Ba levels typically concentrated in basic rocks. These rocks also in pan concentrates when Ce is very high. There are provide a source of metals which are subject to hydrous probably at least two sources of the barium anomalies, oxide precipitation processes, variables such as Mop, firstly, baryte associated with Pb mineralisation and, Znp, Co, and Nis. They generate very strong features on secondly, feldspar or muscovite from the Hendre and geoche m ical maps, due to their presence in peaty Mydrim Shales. As far as is known, baryte mineralistion uplands and areas of active weathering. has not previously been recorded in this area.

11 1 Table 4 Summary of highly significant inter-element correlations (Pearson method) I Correiation coefficients significant at 99.95% confidence level

0.19-0.29 0.30-0.39 0.40-0.49 0.50-0.59 0.60-0.69 >0.70 -- Stream sediments (s) I CU Ass Sns Snp Mop -Zrs Pbs Fes klns C% Pbp Znp Fep Nip Nis Cup Asp 0, Zn, Zrp MnP Pb Nip Asp 3iop Cus Sns Cup Sbp Snp Zns Bas As, Mos Znp PbP Zn CL+ Bas Vs -Zrs Pbp Mnp Cap Pb, Ni, Ass Fes &Ins COS Znp 1 SBp Fep Tip Nip Asp P Ba Zns Bs Ass Asp Pbp Znp Bap Pbs M% Srp -Zrp 1 Fe Srp -Zrp CUS Crs Tip Znp Fep Nip Zns -Zrs Ass Mns Vs Cos Nis

Mn Vs Nip -Zrp Cup Znp Fep Tip -%S Zn, Nis Ass Mnp F% COS co Snp -Zrp Cus Crs Tip Vs Znp Fep Nip Zns -Zrs Ass ?r;lnp Fe, Illn, Nis II Ni Cup Tip +P Cus Mns Crs -Zrp Nip VS co, Fes

V Zns Mns 810s Sns Znp CUS Bs -& Aa Fep Crs Fes Nis Mnp Tip -Zrs Nip I Cr Ass -Znw Fes Cos Bs $10~Nip Nis Vs B Cus Bas Vs Crs -Zrs Nip -Cap Nis Cep -Zrp Up 8 Zr -CU, -Zns -Ass -Cup -Vs -% -Tip Zrp -Mnp -Zvp -Fep -Fe, -C+ -Nis -Bap -Mop -Mnp -Nip As Cu, Bas Crs -Zrs Pbs Vs Mos Znp Zns Nis Fes NIns C% -Zn, Nip Asp 1 MO vs -Znw -Mnp -Tip Pb, Crs Ass Bas Sn cus cup Pk pbs Snp

Waters(w) 1 cu ZnW Zn -Crs -.Ass -Mos cuw 8 Panned coneentmtesIop Znp Tip AInp P cup -Zrs ,\ln Vs 410~ Cup Nip Cu, Zn, Cap Fes Nis -Zrs Znp &In, Cos ;LIop Fep Tip Ti Zn, Ni, V, -310~ Fe, Mn, Cos -Zrs Alop Cap FeP Mnp I -Snp -Cap -Znp -up Ni Pbs Zns AIns hInp Vs Crs Bs Ass Cu, Fes Co, -Zrs Ni, Znp Cep -Zrp Fep Srp Up cuP Sn cl+ -cos cup Gap Pb, Cep SnsPbp II -Tip Up -Fe, As Cus Pbs Zns Bas Ass Pbp Cup Znp Fep 1 Ca Fes -Cep Fes Zn, -S, ;LInpSrp -Up Tip Ce -Tip Nip -Cap Bs -Fes Snp

Sb Cup ZnoL PbS Pb? Sr Zns Bas Fe, Zns Nip Cap 1 Zr -0, Bas -Fes -&Ins -His -Bs Zrs -Cos -Vs -Nip . U -Fep -?Jnp Snp Bs Tip Nip -Cep -Bap CeP I MO Cus Pbs -Zrs Cup Z% Tip Fep hInp Pb? 1

12 B Copper values in sediment and panned concentrate are The rocks in this area are believed to have formed in anomalous at sites SN 2005 3318 (no. 116), SN 2244 3155 an environment which was suitable for the formation of (no. 120) and panned concentrate anomalies, additionally, volcanogenic mineral deposits. A gravity low is centred at others, for example, SN 2260 3229 (no. 129), on the area to the west of Crosswell, over the volcanics SN 2196 3250 (no. 184), SN 2044 2967 (no. 132). The (Cornwell and Cave, in preparation) which might source of these copper anomalies is unclear, though they represent a buried intrusion or feeder to the volcanics. are in part due to contaminants, such as copper wire, The D. murchisoni Beds are of limited extent here and found in the sample from SN 2196 3250. consist of pyritiferous dark shales and mudstones, Another source is the lead zinc mineralisation, Evans suggestive of formation in a sulphurous basin on the (1940) recording chalcopyrite on the dumps. No flanks of the ‘Fishguard Volcano’. On theroretical chalopyrite was noted during the working of the mine grounds, three styles of mineralisation may exist in the (Foster-Smith, 1981). area: massive sulphide deposits, ‘fine’ gold, and vein Au High levels of Assp, Mosp, Zrsp, Cep, Up, Bs and Nisp and sulphides. The latter may be derived from the are also found in this area, the zone of high Mop being remobilisation of metals in the mudstones, perhaps particularly prominent on greyscale maps (Appendix 2). driven by any of the intrusives, or, alternatively, they At SN 1872 3280 (no. 135), the presence of cassiterite may be derived from the intrusives themselves. Although indicates that not all high Sn values were due to the drainage results are inconclusive, they indicate the contamination and that Snsp, Zrsp, Ce , U and Bs highs presence of some metalliferous enrichments and further may reflect a granitic heavy mtner.’ 9 association, work is being carried out to investigate them. particularly as zircon was noted in the mineral examination of some of the panned concentrates, e.g. SN 1834 2798 (no. 144). There is, however, no known Eglwyswrw to the Freni-fawr area local source for such an assemblage. The catchments of the Nant Gafren, the northern tribut- To summarise, anomalies can be attributed to three aries of the Afon Nyfer and the western Afon Dulas sources besides contamination: vein mineralisation, dark cover the Llandeilo-Ashgill succession and contain very shales and granitic heavy minerals of unknown few anomalies unrelated to widespread contamination or provenance. monazite nodules. The area is characterised on the greyscale maps (Appendix 2), by high values of Ce, U, Ni and B and moderately high levels of Zr, Ce and U. Ce and U levels are closely correlated and relate to Brynberian to Crymmych monazite. Some high Ni values are generated by REE This area consists of the catchments of the Afon interference, though some may be enhanced by a high Whitehook (or Bannon), Afon Clyn-maen, Afon natural background in the shales. Bs anomalies (Figure Brynberian and southern Afon Nyfer (or Nevern): These 12) show no relation to any particular lithology and it is drain northwards off rocks of the Fishguard Volcanic not known if they represent a shale or granite heavy Group, overlain by D. murchisoni Shales, in turn, mineral association. overstepped by Hendre shales. The area is characterised The most prominent base metal anomaly is at by Ba, anomalies in the east and scattered base metal SN 1837 3560 (no. 238), where Cup is accompanied by anomalies elsewhere. The spatial distribution of Bas anomalous Zn,, Bs and Ce . Pyrite is present, but anomalies clearly indicates that they are nearly all mineralogical examination fai Ped to identify the mineral derived from the D. murchisoni Beds. High and phase responsible for the copper. The site lies anomalous levels of Basp are recorded from sites downstream of a main road and disused railway, but no downstream of Fron-las (SN 1660 3447, no. 152; contamination was recorded. Prominent Pb, and Pb SN 1614 3546, no. 353) and Fron-lwyd (SN 1730 3482, anomalies at SN 1815 3677 (no. 251) and SN 1340 384g no. 110) and baryte was identified in the pan at (no. 306) are’ shown by mineralogical examination to be SN 1660 3447 (no. 152). This suggests some barium caused by contaminants, mainly lead glass. High levels of mineralisation in the vicinity, perhaps accompanying the Snp are also found at these sites and are probably due weak base metal and possible gold mineralisation tried at mainly to contamination although, at the latter site, these localities. The presence at SN 1730 3482 (no. llO), rutile bearing tin was noted during the mineralogical of AS and Mos anomalies gives support to the possibility examination. The scattered, weak Pb anomalies of a Bark shale-associated gold enrichment in this area. elsewhere in the area are due to monazi Pe nodules. A Base metal anomalies (Pb,, Znp) are recorded weak Cuw anomaly (0.02 ppm) is located at SN 1800 3492 downstream of the Fron-las trials, but with the (no. 237) but its cause is uncertain, although this site is exception of possible galena at SN 1456 3596 (no. 137), downstream of the main A478. A site anomalous for As, mineralogical examination suggested that the anomalies and with a high level of Co, is located at SN 2116 3513, were mainly caused by contaminants. Where examined (no. 181) on the southern slopes of the Freni-fawr. The mineralogically, base metal anomalies from elsewhere in stream is a first order tributary of the Afon Dulas and the area were found to be principally due to the As levels are not high (Ass 55 ppm: Asp below contamination. detection limit), suggesting that hydromorphic processes Pyrite was commonly reported, the sources lying in are the most likely cause of the anomaly. the volcanics and the dark mudstones (SN 1167 3574 (no. 326); SN 1312 3534 (no. 236). The Pb anomaly at SN 1130 3584 (no. 317) probably resu P ts from Pb Eastern Cleddau, Afon Cewgyll and Afon Wern occurring as a minor constituent of rocks forming Many anomalies in this catchment are the result of minerals and weak Pbs anomalies along the Afon hydrous oxide precipitation. Relatively high Fes and Mns Whitehook are probably derived from D. murchisoni values occur over the area (Appendix 2). Locally high Beds. and anomalous levels of Cos, Ass, Zns , Ni , Cu, and High Srp levels are found in three samples draining Znw are also recorded, e.g. at SN 137! 318 (no. 234). the upper reaches of the Afon Clyn-maen (SN 1434 3334, These are most prominently developed in the first order no. 222; SN 1434 3387, no. 223; SN 1430 3445, no. 224). streams which drain the dolerite intrusions. Cap values are moderate, and no obvious local source for The U high at SN 1404 3086 (no. 235) is due to the the Sr can be seen. presence of monazite. The weak Bap anomaly (888 ppm) The volcanics outcropping here are characterised by at SN 1530 3041 (no. 402) is in an area of high Bap high Zr and low Ti , Nip and Bs levels which separate results and is probably derived from feldspathic ashy them quite distinct Py from the overlying mudstones and mudstones of Ashgill age on Foe1 Drych (Evans, 1945). basic intrusions on the greyscale maps (Appendix 2). Large Cusp, Pbsp and Znp anomalies are caused by

13 Lead in Streem Sediment.

. <5lPpm KEY Q 51-70wm -wwmedm . Swnoc* so*

_.___.____D__ 0. 71-110ppm

* rL Mlmh 0. 11 l-624ppm 0. 3 625 ppm

Figure 6 Lead in stream sediknt

14 Lead in Panned Concentrate.

KEY I Q SO-200 ppm Q 201400 ppm

I 0. 401-950 ppm

1 Q 951.1200ppm

. >1200 ppm Cumulattw Froquoncy Plot 0 c

1 a

1 Pmbobblitv

1 Figure 7 Lead in panned concentrate

15 contamination, e.g. at SN 1441 2842 (no. 362), SN 1034 Prominent Asp anomalies at SN 0262 3492 (no. 407), 2853 (no. 363) and SN 1018 2862 (no. 364). Weak Pbs SN 0254 3541 (no. 408) and SN 0226 3424 (no. 427) are anomalies, for example ‘70 ppm at SN 1445 3036 associated with high Cu and, additionally, high Cuw and (no. ZSO), are possibly the result of hydromorphic Znw at no. 407 and z np at no. 408. From element ’ processes, where Pb is fixed by organic matter. Isolated associations, it appears that Cup and Znp anomalies are high values of Vs (242 ppm; SN 1254 2891 (no. 242)) and caused by the abundant Fe-Ti oxides, but the presence of Bs (Figure 12) also occur in these catchments but no the other metal anomalies suggests that sulphide obvious source, other than the shales, can be identified. mineralisation, perhaps associated with the margins of The streams to the north of Maenclochog have only the dolerite intrusions, may also be present in the artificial anomalies, Cusp, Pbsp and Sbp, all being vicinity. ascribed to observed contamination. Bs anomalies (Figure 12) could not be clearly related to any particular feature.

Mynnydd Cflciffeth, Cwaun VaUey and Camingli Sediment geochemistry in this area is dominated by high MynyddDfnas and Newport levels of transition metals derived from dolerite Drainage in this drift-covered area is composed of fairly intrusions. Veti high levels of Mn, Fe and Ti in panned short, northward-flowing streams which cut across the concentrate were shown by mineralogical examination to lithology at right angles. The area is subject to be caused by abundant ilmenite and magnetite. Locally contamination effects from the farms and roads, associated high levels of MO&;i$~d~~~&$$e; resulting in confused geochemical patterns. from the same lithology. All Pb and most Cu and Zn anomalies are caused by associated with the intrusion at Mynnydd Cilciffeth, contamination, perhaps enhanced by abundant timenite molybdenum probably being present in magnetite. Some and magnetite derived from the dolerites. Contaminants of the most anomalous sites are: include: SN 0132 3328 (no. 420), Ti 22%, Mn 1.8%; copper wire at SN 0342 3905 (no.3761, SN 0140 3997 SN 0254 3275 (no. 3991, Ti 24%, Fe 16%, Mn 1.84, MO11 (no.403); ppm; lead glass at SN 0140 3997 (no.4031, SM 9922 3666 SN 0254 3541 (no. 4081, Ti 11%, Fe 2346, (no.412), SN 0230 3875 (no.379); SN 0247 3283 (no. 398), Ti 14%, Fe 13%; various metals at SN 0024 3844 (no.396), SM 9972 3763 SN 0122 3335 (no. 422), Ti 21%, Fe 16%; (no.430), SN 0140 3997 (no.403). S anomalies are attributed to contamination, it is unlikely that these values reflect metalliferous excep“e for that at SN 0228 3792 (no. 3931, for which concentrations of importance, particularly as levels of there is no obvious cause, unless the abundant ilmenite the more valuable metals are generally low (Nip (100 contains minor Sb. ppm, Crs (150 ppm, Vs (200 ppm). Mineralogical examination found evidence of Hydormorphic processes contribute to or cause some Zns mineralisation as well as contamination in two samples and Fe, anomalies over high ground e.g. SN 0744 3290 from the eatchment of the stream between Cam Slanney (No. 273). The highest Mns levels are located on lower and Mynydd Dinas. At SM 9972 3763 (no. 430), a little ground, e.g. SN 0410 3452 (no. 340) and SN 0222 3358 pyrite and a gold flake accompanied heavy (no. 419). Nis, Crs and Vs show iregular distribution contamination. Composition of the gold, which XRF patterns (Appendix 2) with locally high values, for analysis found to contain Ag but not Cu, Pb OF Zn, example Crs at SN 0588 3454 (no. 288) and high levels of suggests that a natural origin is possible. Upstream at Ni,, influenced by secondary processes, accompanying SM 9922 3666 (no. 412), a grain of chalcopyrite was Mns at SN 0410 3452 (no. 340). A very prominent zone of identified among the contaminants, Acid volcanics and high Cuw and Zn w results characterises sites draining dolerites form the bedrock lithologies in this catchment. the southern slopes of Carningli, for example at SN 0518 Pyrite was found in several concentrates, for example 3606 (no* 436), where Zn, and Cuw values of 0.08 ppm SN 0140 3997 (no. 403), but may simply derive from the and 0.03 ppm, respectively, were recorded. The source of small amounts found in most rocks of this area. the high values is unknown, dolerite intrusions on high Bedrock generates anomalies locally, high levels of ground elsewhere failing: to generate such anomalies. Zr deriving from acid volcanics, for example SN 0220 There may be a relationship with some other related 3811 (no. 410) and SN 0222 3786 (no. 411). Anomalous factor such as the presence of sulphides in or marginal to Bas (1290 ppm) at SN 0024 3844 (no. 396) is also ascribed the intrusions. to this source. Dolerite intrusions generate high and Many base-metal anomalies are caused by con- anomalous levels of Tip, Mnp, Cap, Fep Mop and other tamination due to the proximity of farm buildings and elements found in ilmenite and magnetite. The most roads to the sample sites. Lead glass, found in samples at anomalous sites are: SN 0374 3734 (no. 3851, Ti 12%, SN 0218 3430 (no. 426) and SN 0660 3416 (no. 265) is the Fep 1146, MO 5 ppm, Cap 2.5%. and SN 034! 3772 commonest cause of Pb anomalies. Contamination was (no.3 69), Tip &, Cap 2.6%, Srp 19; ppm, MO 4 ppm. noted at SN 0629 342% (no. 266) and SN 0577 3338 Locally, hydrous oxide precipitates pr 08uce weakly (no. 285) but the source for the Pb at SN 0444 3499 ‘elevated levels of several elements derived from the (no. 327) remains in doubt; miner af ogical examination same source, e.g. SN 0430 3796 (no. 384) Co, 47 ppm, identified lead secondary minerals of uncertain origin Fes 6.9%. At SN 0140 3997 (no.403), high Crs, Vs, Cap and no contamination was recorded at site. No and MOPlevels are derived from Llandovery sedimentary contamination was seen at the sites with anomalies rocks, intruded by dolerite. (SN 0223 3479 no. 406 and SN 0218 3430 no.=I 4 6) but it is likely that they are also derived from unseen contaminants. Pbs anomalies are weak ((70 ppm) and are Afon Clydach to Brynberian considered to be caused by contamination and organo- This area is dominated by outcrops of the Fishguard metallic complexing. Volcanic Group, which give rise to prominent Most Cu anomalies are likely to be caused by small geochemical features on the spatial distribution plots amounts of E u in ilmenite and magnetite, as they show a (Appendix 2). The most notable is that Zr levels are high strong correlation with Pep and Tip. The source of the over the acid volcanic rocks (zircon was noted in the anomaly at SN 0322 3397 (no. 437) was not identified but concentrate at SN 0983 3705 (no. 300)). Other features is most probably contamination. High and anomalous Znp related to the volcanics include low levels of B,, Tip and values have a similar source. Nip In strong contrast, dolerite intrusions are

16 Zinc in Stream Sediment.

< 371 ppm KEY . 37 l-480 ppm 0 - n@orrMdl

Rgure 8 Zinc in stream sediment

17 Zinc in Panned Concentrate.

. < 250 ppm

0 250-301 ppm

0. 302-440 ppm

0. 441-614 ppm 0. 3615ppm

Figure 9 Zinc in panned concentrate II Barium in Stream Sediment. < 900 pm KEY 1 Q . 901.1399PW 1 0. 1400-3099pm 0. 3100-3499PPm I 0. a3500PPm 8 1 I 1

8

t I 1 I 8 8 1

I Figure 10 Barium in stream sediment t 1 19 Tin in Panned Concentrate.

. ClOppm KEY

0 10-l 38 ppm

0 139-254 ppm

0. 255434 ppm

0. 435-619 ppm 0. 3 620 pm

Figure 11 Tin in panned concentrate

20 Boron in Stream Sediment. I . <107ppm

0. 107-113 ppm 0. 114-120pm

1 . >120rqxn 0

I Figure 12 Boron in stream sediment

21 ._. .

Figure 13 Anomakus areas described in the text responsibl for the anomabs in the extreme south-west Cilciffeth or from contamination. M at SM 9946 3274 of the area, where geochemid patterns reflect the (no, 4911, may’iimilarily relate to the% asic intrusion and lithology and are similar to those of the Gwaun area. cos and Nis anomalies could be caused by hydrous oxide High 2% derived from ilmenite is found at SN 0714 3595 precipitates from the same source. Anomalous levels ,of (no.3311 and SN 0712 3610 (no. 332), whilst an M high barium in sediment and panned concentrate at at SN 0791 3555 (no. 298) is probably derived ?pram a SM 9696 3140 (no. 463) and barium in panned concentrate similar source, at SM 97323215 (no. 465) were recorded. In both cases, P and Sb anomalies, for exampie at SN 1122 3697 Hap values are higher than Ras values and baryte could (no. 34 4:) and SN 1056 3642 (no. 3501, are caused by be present. An alternative source could be in the Arenig contamination; lead gU and various metals being the mudstones, intruded by dolerites, in the catchment. A tin main causes. The Znp anomaly at SN 0983 3705 (no.390) in sediment anomaly at the latter site is probably the is also caused by metallic contaminants though Fe-Ti result of contamination. Pbs anomalies may be due to oxides may aiso contribute. hydromorphic processes, e.g. at SM.9696 3140 (no..463), The only evidence of mineralisation in this area came but contaminants are a more likely source. Pbp and Sn from grains observed during field examinations of panned anomalies are also mainly caused by contaminatio R, concentrates: a flake of gold found in the panned though possible cassiterite was recorded at SM 9859 3476 concentrate at SN 0983 3705 (no. 3901, near the (no. 4491. Neolithic cromlech of Pentre Ifan, and a possible grain P.yrite is recorded at SM 9695 3554 (no. 4461, of chalcopyrite noted the pan concentrate at although the Cup anomaly is thought to be derived SN 0752 3836 (no. 361); A recent gravity surveys have primarily from metailic contaminants. reported the presence of a negative anomaly in’ this area, There is little positive evidence for mineralisation in which might be related to a buried intrusion, a vent, or this area, but drift cover, low sample density and thickening of the volcanic pile (Cornwell and Cave, in contamination. would have made it difficult to detect. preparation). No evidence was found from the drainage . 2 survey to suggest that if a vent or intrusion is present it I ’ has any mineralisation associated with it. The majority of anomalies in these catchments are caused by basic intrusions in the Arenig and Llanvirn Crinef Brook, Afon Cledduu, Naby-bugail shales and mudstones. Sites at SN 0181 3170 (no. 476) The streams of this area flow in large, deep channels, and SN 0177 3168 (no. ,477) have Mnp and Tip anomalies which havs’been cut sub-glacially and range in size from derived from the Mynydd Cilciffeth dolerite. The Mnp the marshy ditches which are the headwaters of the and Ni anomaly at SN 0221 2904 (no. 489) is similarily Cleddau to deep, fast-flowing streams such the Nant-y- relat J to basic intrusions; tin in contaminants as well as bugail. Where not well drained, the ‘bottom’ land is wet cassiterite was noted in the concentrate from this site. and marshy. In the gorge of the upper Nant-y-bugail and The Pbp anomaly here is caused by contamination from River Aer, m~uchexcavation and infilling has taken place Pb metal and lead glass, whilst Cup may be derived from to accomodate the Royal Naval Armament Depot, contaminants or Fe-Ti oxides, whose source lies in the Trecwn. At site SM 9899 3424 (no. 4901, in the depot, dolerites. many elements are anomalous; Fep, Tip, Mnp, Mop, Cus, High and anomalous levels “if &us, As, and Co, at Pbs, Nis and Mns. These could be partiy relate$l to -P&Ti SN 0340 3010 (no. 478) and Mns ayd Cos at SM 9845 2907 oxides derived from the basic intrusion at Mynnydd (no. 468) are probably caused by hydrous oxide

22 1

precipitation and the presence of basic intrusions. Other RGFRRENCES base metal anomalies are attributable to contamination, I e.g. SN 0202 3060 (no. 4?4), anomalous for &@-and&& BEVIWR. E. 1978. Pumpellyite-bearing basic igneous which lies downstream of a road and disused railway 6X &ocks from the Lower Ordovician of North trackbed. The weak Cu \. ,# anomaly’ ‘161 ppm) at , ’ SN 0117 2989 (no. 473) is !&her downstream and is,..&* Pembrokeshire, Wales. MineraL Msg. Vol. 42, -‘w.~-,-pp. 81-83. 9 8 accompanied by a moderate Sn level of 65 ppm;possible : -and ROACH, R. A; 1979a. Early ‘Ordovician- pyrite was noted in the concen Prate at site. The mineral i volcanism in Dyfed, South Wales. in Harr& A. L., responsible for the Cup anomaly at SM 9932 2977 \. Hollahd, C. Hi., Leake, B. E. teds). The Caledonides (no. 469) was not represented in the” dense (SG >3.3) fraction, examined mineralogically. Contaminants and . 9 . of the British Isles. GeoL Sot. London. pp. 603-609. 8 - 1979b. Pillow lava and isolated pillow breccia of abundant ilmenite are both-present in the conqentrate. ” Rhyodacitic composition from the Fishguard Volcanic The source of the Bs anomaljr at SN 0284 3Oi4 (ijo.484) -’ * ’ i is not known. Group, Lower Ordovician, SW Wales, United -.* Kingdom. I Journal of Geology , Vol. 87, pp. 1937201. t r’ J T -- . . * Afon Syfynwy BRADLEY, R. I. 1976. Soils in Dyfed III: sheet SN 13 Anomalies in the headwaters of the Syfynwy show (Egiwyswrw). soil Surv. Rep., 38. precipitate associations, but the*hAmain source ef+ -the%“& -I RUDEF~RTH,‘C. c. md wmm, c. 1977. 8 anomalies appears to be the basic intrusiv& and, the, Distribution of some chemical elements in the mudstones which crop out in the catchment. High levels ’ soils of North West Pembrokeshire. J. soil Sci. Vol. of Cr, are derived from these lithologies, as are V,, Fes 29, pp. 258-270. and Mns anomalies. Ilmenite and magnetite from the CHARLESWORTH, J. K. 1929. The South Wales End I basic intrusions are suggested as the sources for the ’ Moraine. Q. Jl. Geotl. Sot. Land. Vol. 85 pp. 335-338. Pb Zn and partly for the Cup anomalies at COOPER, Di C., NUTT, M. J. C. and MORGAN, D. J. SN%14 2!60 (no. 482), although much of the Cu is likely 1982. A reconnaissance geochemical survey of to be contained in contaminants. Contaminants’ may also Anglesey. Mineral Recomraissance programme Be@. I cause the Sn anomalies in the catchment. Arenig age I@. GeaL Sci. No. 51. . shales and g ealyham Volcanics probably - cause the -, BASHAM, L R. and SMITH, T. K. 1983. On the anomalies at SN 0830 3160 (no. 282) and SN 0825 3061 occurrence of an unusual form of monazite in panned (no. 284), wh ere Pbs, CrS, Mns and Ass are high. stream sediments in Wales. GeoL J., Vol, 18, pp. 121- 8 The source of the Bas anomaly at SN 0659 2822 127. (no. 481) is not known but it could derive from the COOPER, D. C., CAMERON, D. G:, 1 -1. Arenig shales of the locality. JACKSON, AUDREY A., ALLEN, P. M. and PARKER, M. E. 1983. Exploration for volcanogenic sulphides at Benglog, North Wales; .-Mineral 8 C^dNCL,irsrok$ AND RECOMMBNDATIC$~~ * s ’ - ‘,= -:” R-Programme Rept. hst. Gee& Sci. Most’ of’ the significant geochemical var@ti&: could $2’ , re’tited’% .bedrock lithology (particnlariy ~d’dol&rite &d ,-No.63. .; : * mudstones); to contaminanation; I known *mirf&$lisation e CORNWELL, J. C. and CAVE R. in.preparation. An I and to hydrous oxide precipitat&n a&l scavenging. Airborne Geophysical Survey of *part of Dyfed Two areas” where anomalies &g&ted some minera Mineral Reconnaissance Programme Rept. Irrst. GeoL potential-are considered worthy 'of more deyled study: : ‘:. Sci, . . c . COWPER REED, F. R. 1885. The geology of the country 1. ‘%a, Cu, Pb’ and Zn anomalies in catchmerits draining’ .,-around Fishguard, Pembrokeshire. Q; J. GeoL Sot. I ._the : -Fishguard ‘Volcanic-z~Group and overlying D. a. London, Vol. 51, pp. 149-195. . COX, A. H. 1930. Preliminary note on the geological ’ Mukhisoni Beds: in the Crosswell-Crymmych area: structure of Pen Caer and Strum.ble Head, The presence local& of As enrichments, old trials and’ Pembrokeshire. Proc. GeoL Assoc. Vol. 41, pp. 2740 . records of possible~gold mineraIis8tion also’suggests .. .* I < ‘some ,potential” for gold :mineralisation associated 289. : with ’ the dark mudstones. Volcanogenic massive ELSDEN, J. V. 1905. dn the ignebs ro& occuriring . between St. David’s Head and Strumble Head, sulphides, gold, and vein sulphides areconsidered to Pembrokeshire, Q.J. GeoL Sot. Lo&. Vol. 61, be the most likely styles of mineralisation. Further 8 work to investigate these possibilities is in progre&s. pp. 579-607. EVANS, D. C. 1906. The Ordovician rocks of Western ,. Carmarthenshire. Q. 6.. Ged. e. Land. Vol. 6+ ( * 2 Pb, Zn and Ba anomalies ‘in .the Hendre and Mydrim , ishales of the Taf Valley. An extension- of the drainage pp. 597-643. - . 8 survey is required to delineate the boundaries of the‘ EVANS,* W. D. .1938. The Arcnig; LlanviA and ‘Lo.wer anomalous area, !- and more detailed geological; Llandeilo rocks of the Eastern Prescelly Hills, geophysical and geochemical studies are needed to Pembrokeshire and the eastward termination of the. locate the source of the anomalies and styles of Fishguard Newport Volcanic group. PhD the&. Univ. mineralisation present. *. .. ** W&s. (unpublished). I - 1940. The geology of the Prescelly Hiw’and adjoining In addition, the following are tenuous indications of mineralisationt - c. -c. areas in North Pembrokeshire and Carmarthenshire. Ph.D thesis, Univ. London. (unpublished). . a Grains of .chalc+yr&e and gold in streams - 1945 .The -Geology of the Prescelly Hills, North 8 draining Fishguard” Volcanic Group rocks and Pembrokeshire. Q. J. GeoL Soc. a_,Vol. 101,. dolerite,,near Dinas’and, Pentre Zfan. -’ pp. 89-107. ’ - b Scatter&: ’ occur&es ‘t of cassiterite grains, FOSTER-SMITH, J. R. 1981; The.Non-Ferrous Metal :, Mines of the South Wales Area. British Miniqg. always only one or two and bearing no obvious .I I ‘vol. I$. 8 relationship to the geology. _ Imperial CoR&e of Science and Technolgy, Applied ’ l Geochemistry Research Group 1978. The Woo&n “:‘ ACKNOWLEDGEMENTS z Geochemicd Atals of Rnglandand WaIti; (Oxford.“~,~ . 8 Sample collection was undertaken in part by Mr P. Bide. Clarendon Press) 6 9p. All samples were ,prepard and analysed by staff of HALL, G1 W. 1971. Metal Mines of Southern kales. * Analytical Chemistry Unit, BGSi _ (Gloucester: John Jennings Ltd). ,t ,7’-, ,_ !,L‘ ’ ..;

I 23 .I JQHN, $5. s1196S. A Poss&le Main Wurm Glaciation in West Per&wire. I&ure. Voi. 207, pp. 622-633. - 1967. Further Evidence for a Middle Wiirm lnsterstadlal and 8 M& Wiirm Glaciation of South- West W&es. GeeI. IUq. Vol. 104, pp. 638-633. - 1976. Pembroke&&%. In C. A. LEWIS led), The GEsciatBb3lrs~WalnrsaadthcAd@i!k@aegioRa (Londont Longmsns), pp. 239-265. - %972.The Fishguard-Pembrokeshire Area: a description of the O.S. one inch sheets 138/151: Fishguard and Pembroke. British Landscapes Thrwgh m Vol, 16. JONES, 0, T. 1912. The Geological structure of Central Wales and the &joining regions. Q. 3. Geol. Sot. Lo- Vol. 98, pp. 328 - 1965. The glacial and postglacial history of the Lower Teifi valley. Q. J. GeoL Sot. London. Vol. 121, pp. 247-281. LEPELTIER, C. 89629.A simplified statistical treatment of geochemical data by graphical representation. &m. &oL Vol. 64, pp. 538-650. LOWMAN, R. D. W. 1977. The geology, petrology and geochemistry of an area of Lower Palaeozoic rocks east of Fishgu&d, nurthern Pembrokeshire. Ph.D. u&58& Univ. wale% (unpublished). -and BLOXHAM, T. WC1981. The petrology of the Lower Palaeozoie Fishguard Volcanic Group and associated rocks east of Fishguard, N, Pembrokeshire, (Dyfed), South Wales. J. GeoL Sot. Lo&em. Vol. 138, pp. 47-68. NOWLAN, G. 1976. Concretim mangeon oxides in streams and their usefuti as a sample medium for geochemical prospecting. JL Geocbem. Ex@or. Vol. 6, pp. 193-210. PARSLOW, G. R. 1974. Determination of background and threshold in exploration geochemistry. JL Geochem. Rxplcr. Vol. 3, pp. 319-336. PART, G. M. 1922. Notes on the Ordovician Lavas Of the Mynydd Prescelly, North Pembs. GeoL Ma& Vol. 54, pp. 318-323. READ, D. Rare earth element behaviour during subaerial weathering. PhD thesis Univ. Lo&u (unpublished). REEDMAN, J. H. 1979. Techniques in mineral exploration. (London: Applied Science), pp 533. SINCLAIR, A. J. 1976. Applications of probability graphs in mineral exploration. Assoc. Explor. Geochemists Spec. Vol. 4, (Canada. Richmond Printers, B.C.) 95p. THOMAS, G. E. and THOMAS, T. M. 1956. The volcanic rocks of the area between Fishguard and Strumble Head, Pembrokeshire. Q. J. GeoL Sot. a. Vol. 112, pp. 291-314. THOMAS, H. H. and COX, A. H. 1924. The volcanic series of Trefgarn, Roth and Ambleston, (Pembrokeshire). Q* J. GeoL Sot. London+ Vol. 80, pp. 520-548. URQUIDDI-RARRAU, F. 1973. Regional Geochemical Variations related to base metal mineralisation in Wales. Ph.D. Th&s, Univ. London. WILLIAMS, T. G. 1933. The Precambrian and Lower Palaeozoic rocks of the eastern end of the St. David’s Precambrian area, Pembrokeshire. Q. 3. Gee%.SW. London. VoI. 90, pp. 32-75.

24 w xaf 2so.14

f::% r1o:to 200.00

: t :*:I 21o:oo

6;;;.:;

1~5o:oo 5110.00

“::1*88 Il4o:oo PIG2 3 2NP XR7 FtP 'XRF NIP :RF PLP7CXBi 73.00 6:50

::z 26.00 55o:oo

‘::-:: :x so:00 ;z*:i Blo:oo 105:00 ::-ii :M ro73:oo 85o:oo

299.00 X8% 7755o:oo

41.00 631.00

215.00

::*xx 550.00 23:oo WiA8 3e.00 362.00 27;;Ai 75o:oo :xs 450.00 96:OO 629:OO

‘W8 355:oo t:::M 7139o:oo %*81 13:50

117.00 t::+ 640.0: t::* 72:00 ‘::-8t 10610.00 161: 2:.00 73:ao W! 2;tw;.;i 120.00 [email protected] 6ZO:OO tt:: 47.bu :::II: 254.00 101910.00 3470.06 3232o:oo %iX PXCL I5 PCP XRF ZNP XRF 8AP XRF FtP XRF IIP XRF l;:W;~.;;

97070:00 l7040.00 10239C.00 1190.00 13.50 13::;.g ::-:i 16C.00 29:CQ 13.50 s7c:cc 130.00 79.00 370.00 ‘f;.:: 1g.g

b:5Q 29C:CC 70.00 91040.00 53.00 I:;:;~.::

::+ I369C:CC 39:cc 1p; ‘:;~A8

II:00 72C:CC 29.00 330.00 2;g.;; ml: S&:00 4Ic:cc

63.00

:::::*I: Is4Cc:cc lC76CC.c~ 77000 .QC

:i:::*xfJ ss7~l:cc I6 IN? llr 4.50 4.50

::*:: IO:00

w: 445:09 2j.40 20I5.00 ';:;;A: 'X r:ro 1220:00

:;;l.;{ 7ooo:oo l:;m;.~f 1120.00 5540:00

1000.00 '::*:I 241.00 r:so ':f:*l: 14.00 9oo:oo

':*:: ':::*:I 4150 241o:oo 1420.00

I:::!: 445:oo

':::*:I 14ll:oo 14200.00 tltf. 27 1tlP XRT IIP XRF ::% BAp13x:i ::r,oxaa 46910.00 1;;:;; 5406o:oo

110:00#:t8 13:50 :::9w 13.50 r52lo:oo 110000.00 1:t.t: W81 103690.00 f;:::; 31:oo 11c100.00 1t;;:;.;; 90:oo “:M 102.00 99:oo 90610:00 138.00 f:I*:! W81 ‘:K 5;::s”; B469o:oo 1oL:oo 0269.00 2:;:g mf.88 ‘X X-81 160.008l:oo 2;g:g 024lO:OO 118O:OO 104140.00 1450.00 Ill:00 141570.00 w;.# :0:+ -1:oo ‘Y:Ili ““‘kI18 -1:oo lr anr sNf5“:: ‘l’ 1% Yoo”:! “O’ 44:: 4:50 I:01 4oo:oI ‘:*:I I”{.:: 1.00 t;::: riir I:.:: ::f:*:l I.IE pi 155c:ol 1o:oo ;:!I :.!I t pi 1:oo yi:ti /:I! 5:o I:::! 155:04 I:10 :*:I :*I! 1:tc 4:50 1:oo JMl ‘:*:I b.00 1:;; i*:o ::3 I:.:; i(:i;:ii 1:oo 4:!0 I:10 ?.OO ::g 210:00 11.;: I.00 :*iII 4150.00 4.40 4:54 1:oo pi 4.50 ‘;.I: ::::*:I :*I: ;::::I: 1po” $110” ‘!:*:I 4:oo 4:50 5.00 5y::: 1:10 1:oo -1.00 -1.00 -1.03 -1.00 S

1420.01 mm 23171: W8 290:00 :::::* 23449:

:::::* 21x0: 21331.

:t:::* 2137t’ ::-ii 21066. 56:00

:t:::* 21306: 62...OO :w 599.00 21291: 21119. 2lI91.

:t:::* 21295: x:* 21948:

t1s02.

:tf::- 21370:

:w 23149. 21400: 2244r. 23?69. 23431. 23334, 23397. 6 cR;200;; I Not3”:: WC 2O:: 136:OO 0:50 2:50 1.10 2.50 ::+ los:oo xi 3: p; 2:5b

I:tr

YO1.00 ta:: I:10 :3:*8t t:a 0.50 19o:oo g::; 245.00 :% 191.00 $88 pi J55.00 W8 :p; $38 76:OO 19:oo 1:lJo I::.:~ ':% :*:i 02.00 693.00 r:oo I:20

:*:i OO:;;

;::;

'::*llt 1:;; 72:00 35.00 p;

2:oo

:::.ts ta:: 206:OO 1:40 2.30 :::%I 329:oo :*:I 4::;

I:10 70.00 2.10 61.00 12.00 1.90

'!*1! ?!I pi I:10

II:00 t*!8 1:oo

xi a:50

E-8: :i*Ei 225.00 :*fi: 9s:oo 46:OO 246.00 I:70 104.90

129.00 91 .oo

%*8! 10.00 t;;::O” ::*:i 10,:oo Jr:?0 1110.00 41.0; i::.:: 42.00 150:00 ::t:*:: 606.00 910:00 :1*;! 24:oo ::;*I: t:::.:: t::::*:: 13:so 94u:oo 12230:oo 1210.00

;::*:: 11:oo

t::*:i 159:oo

K0*8! 16560.00 13fo:oo :::1;:+ r949:oo lb249.99 I7

21056. 1000.00 2p5.g ::::t* 21614: 191:oo 545.00 :::::- Wl! 22911: 699:00 :::::* 21154: 21441. 22142. :ti::- 611.00 2P946: 20149. M-81 20341. 794:oo 54D.00 :oi::* 20968:

1120.00 #X-XI 188:oo 937.00

‘::!*!t 2520:oo ::f:*li r19:oo 'M 732:OO :::1*:8 446:OO 'Z*III :8x* 13oo:oo 20247: :wt :::::. 64O:OO 21469: 1060.00 CUP XIF

12.00

‘:-I: 16:OO

‘:*I: 13:oo

Y-811 3:oo

:::::

1321 .oo 3311.00 44.00

:::%8: I:;;;:{;

:;;;I::

1213:oo

5356.?0

‘“E*:! ‘;:~;:g

200. 46O:OO 201. IM! 7r7o:oo WI :::* ::i 491:oo 204: 430 1s900.00 w: 21o:oo ::+ ::x~ 240.00 ‘$9; “:$P8 ‘ETJ 4:so g:g 42O:OO 4.50 600.00 r4:oo 1173.00

2mg.g:

IO.00 35o:oo 1257O.tJ ‘:.E pg tx+t 14990:00 4:50 22290.00 tw 1410:o0

124O.L’J

‘::!*:I Il. :oo

1~400.00 PIG’ 20 ASP XRF CAP XRF CEP XRi ! iBF XRF UP XRF I.00 0.00 :*z ‘:TJ 4:oo 1:10

t*tIl I:10 :wo” 146o:oo

:wt 4.00 1190:00 5.:; 7940.00 Iv-i.?” 1:oo Y9lO.00

:*I 1:oo

‘X Ior: 420.00

16900.00 2wg.g:

216O:OO 340.00

::w W8 2510:oo soo:oo

1121.00

‘“X8 W8 1209:00 71o:oo :-s: ::E! 5g,g 2:oo 47oo:oo 2140.00 6r:OO %: 1:oo w: 1.00 86l:Oe 1109.00

:*tII 3410.00 t:l*!: 3:oo 6410.00 1ao:oo YNC OIS 1i10;00 ::::.:I llYo:oo

:::I% 51ro:to

1w:.t4

1to:oo 0414 DESCRIPIION ______rllE Tl!?E :IEYPf ILE

NO. Of f1ElDS t 40 NO. Of ltCORO5 I 350 WORDS P[R RECORD I 40

C4RD INPIll FCRYI!

.I.. . ..I ..I. IWE FORM41 USED TO PRINI IWE fOURlH AND sUBSEPUEN1 LINES IS IHAT USED FOR 1HE 1HlRO LINE...... If MAY PRCDUEE ODD RtSULtS. IN WIICH C4Jt USE GPFlNlWltN 4 SUIlABLL FORYAI 10 PRlNl ;HE *.a. ..*. FILE. IHE DATA WILL BE bNH4RMED. 4llliOUGH It YAV LOOK ODD. ,......

SAYPNUYB ElSllNt NORlNlNG cut 44s Pat 44s 1NC 445 RAC OES fEC OES HIiC OES cot OLS __~_..______~~~~~.~~~~-~-----~~-~..~~~~~.~~~~~~~.~~~------~~~~~~~**~*_.*...... _..___.___*__...*_ flEL3______LENCIH tfm;:PL *. , . . . . , . . . . * . . . . , *. .a 1 . . .* , . . . . , . . . . 1 , . . . , . . . .

mm*....*.. f f f f f f r f f :PPIR LlYll ---.******* 491. 22260. 23991. 60.00 4100.00 1100.00 5190.00 9IPOO.OP 25700.00 261 .OO LMR Lllllt *__..__..._ ;k 19430. 22799. 5.00 20.00 30.00 197.00 13200.00 124.00 5.00 4lStliI VALUE ..*l.-.**.**....* DlCtlONd’~t6’t~Nt l&itlFllR 0. 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ...... __..___.___._._*_....

SAYPNUYS NIC OES VC ors CRC CiS It OES 2RC OES ASC XRF YOC OES SNC OtS clw 44s ~~~..~~..~..~~~.~~~.-----.------.------~.~~..~~~------~~~~~.~*~~~.~~~~~~_.~~.~~~~~.~. ______...__._ FtELD___..__.__._ LtHCIH FtELD .TVPE . . . 1 1. a* , .I .I , *(I .I 1 . . *. * . . .I , .* . . , .a *a , . . . . * . . . . _____.___. f f f f f f f f f tPPER LlYll -...... m. LbwR LlYll491. 129.00 242.00 220.00 131.00 4230.00 90.00 19.00 106.00 0.11 _._._____._ 4lStNl D414101. VALUE 5.00 14.00 21.00 ll.Ob 117.00 2.00 o.so 2.50 0.00 ...... *..... !3lCllOH4i:~SECVtN7 !dHllflCR -1.00 -1.00 .I.38 -1.00 -1.00 -1.00 -1.00 -1.00 ------~~~~~.*...f~._.**....

‘~UPNUMB tuu 44s Fil? XRF PBP XRF INP XRf B4P XRf fCP XRF HNP XRf \IP XRF NIP XRF ..~...~~~~~..~__----.----~.~..-.--.....--.....-..-.-~~~~~~~...... ~...... ~..~.~...... ~....*...... ~~.._.~__~~~~.~..~.~ i’lELD LfhCIH

-_-._.-w.. f I f f f f t f LPPfR LlYll ___..._____ 491. Yl40.00 2li9l4.10 99lS.00 14200.00 234510.00 18t40.00 241000.00 lO~.OO LC%IR_.-._-_.... LIYII AlSlNl 0414101. VAlUt 0.00 3.00 6.50 24.00 11.50 13690.00 210.00 2310.00 2.50

-1.00 -1.00 .I.00 .I.00 -1.00 -1.00 -1.00 -1.00 DIC7lON4i~~St6Y~N7*~EPallTI[R

SAYPNUWD ...... ~Y’..~~~....~~~~*~~~....~~~_.~~~.~..~~~__~~~.~_.~~~~_~~~~~__~~~__~~~____~~~~.~~~~~~.~~_~~~~~~___~~~_~?~~____ tlELD LtNGtH ______tfm;:rr .* 1 .I I. : . . . . I . . . . , . . ** * a. I. , .a 1. , 6. *. , .I 0. , *a *. _.__.*.___ r f f f f f t f f UPPER LIII~I ._._.._.__. 491. 2435.00 66.00 44140.00 14500.05 111.00 100.00 4210.00 101.00 19.00 LcMlR LI~III _ _ _ _ . . . _ _ 101. 4.10 1.00 200.00 Yf.00 4.00 1.00 90.00 I.10 1.00 AbSlNt 0414 VllUE ...... -..___..._ -I 00 -1.00 -I .oo -1.00 -1.00 -1.00 -1.00 -1.00 DIC1ION4~:~SltY~N7-f~:~llflEX 8C OE5 45.00

lp8 289.00

6l:OO 118.00

60.00

:*:i 242. 2:50 2.50

‘!*S! 8.00 2:50

249. 68.00 251.

13% II:00

If:.:! 561.00 tro:d

x0” 15:oo 157.00 15.00 El! 12:oo 78.00 ::;* 211: ::.o”: 289.00 4o:oo 10.00 %*PS :i*z ZG5.00 ho:00

3:: 73:oo 40.00

:::I: ::%i 64.00 f:8::! 46:OO PlCE 11 ftP x11 TIP XRF Nl? XRf *NLSX~; 61640.00 25.00 164:OO 8!110.00 ::::1% 1326O:OO t::+e” 152:oo 1270.00 ;I::x 15s70:00

XII-:! 12110:00

31.00 199.00

‘;:*I :i;::: 6:50 ‘z-9: ::::x I,:50 11910:00 It;*88 30010.00 13:50 ;:;#A8

:::- 211.00 6SllO:OO 1220.00 261: 70410.~0 1020.00 ‘::*I!13:50 9846b.OP 1010.00 13?250.00 t?g*;; 191400.00 96.00 124200.00 2g.j; 110610.00 j410:oo ‘;:;;~A; J6l:Ob 9229b:OO

x:*:8 13~~o:oo X# x1-8: 19,10:00 98tlo:oo 102610.00 114010.00 lo7oIo.co

51.00 398.00 506.00 :::*I 42O:OJ ::8*f! :::- “;:A! 140:00 255: 19’00 :::a 7ro:oo 3 I;$.:; 258: 17o:oo :::* 261: 2b2. 10.00 21.00

260. 269. 210. lb.00 21.00 ‘:-:o” 4150 :.:i 4:so ac OfS Il.00 lioc0”:: y$ 1:00 ;p; I% p; 7u:oa 58.00 a:50 Il.00 0.50 31.00 80.00 I*:: g;*g 1:oa 0.50 a4:oo :*:i ;:PJ 410. 0:50 :;:* 422: I% 1:ro :::* ljg.81 425: 0:50I 3 114:oo ?l.OJ :::- 104.oo :;.g; 428: ::*ie 1g.g s-:x 429. 7s:ao p; 2:so 430. 90.60 a4:oa 4i.g 411. 66.OC po” 16.00 ::-!I 2:u 65.00 ::*z 39:oo pi ‘“:A; 101.00 5S:ob ‘;:A~ 121.00 ::*:i a:50 230 16:OO 0.50 99:oo kx 8::: p$ 13.00 0:50 bC.00 I:;.$ !*:II p; 71:ot pi ::*io 1s:oo a:50 61.00 i-:0 1:20 ‘:A; x.:: 2:so 1:so 1% 0:so CUP XRF PIP XRF 21.00 33.00

‘t:% I:00 2060.00

44.00

0.01 ‘:!% ::x! 6O:OO 56O:OO 1% 16.00 0:oo 22.00 t::z 1,5:00 Z.18 9:oo 14.00

‘Wo” 2;ht X% ‘::!t 0:01 14:oo 502o:oo 31.00 6120.00

‘::G 49:oo ,f~# a:00 Il.00

::*t: 13:oo

:::. ;:-I! 454: x:00 456. -.s: 2130 b.:O ZQ.Cl ws :::*461: ‘1.14 14.00 l97o:oo 25.0’ 19 .>a %11*88 I’,.39 236O:OO 16.00 2150.00

l.& :t*tk! 9:oo :::: :::: 14.00 24 SIYPNUYI 402.

:::. 405:

:t:. 408: 409. W18 1104:00 ~115.00

110.00 490.00 31.00

21.0

‘33 cl:00 11.00 ‘;:A;

19:oo

“:% 520.00 4:50 ‘;;w~.~;

49:oo 110.00 546.00 1924.00

‘X 4:E.o

461. 462. :::i 1lr:oo

::*8: 4150 %88 54.00 410:00 :::*o”: 75t:oo ::t::* 21671: 3Cl.00 ‘X-88 t::. :t:::* 915:oo 111: 21936: :::::- 23696: :w 121:oo

624.05 16.00

:x 17:oo 152.00 :::::* 397.00 ‘X 22143: ElmIX 75:oo %!I 57Q:oQ E8:88 :t::t* 40:OG ::*1: 21911: wll 11:oo :8:88 1250:OO %X: ‘:*:I 2x54. :8*ts ::x IQ:00 3o:oo 706o:OQ

mu;.PJ

1010:00

5QO.00 f::%E :1::: s11:00 iE8.88 751:oo ::::81 22191. 71:.40 23164. %131 191:.GO 529:01 15.00 537.40 :1:8*81 9os:oo *::1+ Mm;.;; 539:oo 766:OO 5.40 152. tE.8: 900.00 151. 45f:oQ i:%i 161. glg.83 23:oo :t;*lli 1170:00 22,09 it:: 37o:oo 11300.00 111.00 2 sIYPNUY1 VC OES IC OES 101. 115.oc ::%l E% 11:oo p;:;; 14.00 lil:OO :I*88 15:oo ::+ 15.00 92:OO 114.10

f::*IX 192:OO :x 211.00 16:OO t::+ 16.00 91:oo

':*:t 0:50

99.00 59.00

'ii 9: 6l.D" 'y: ii'!! 94:oo ri:oa ;:t% 59.00 115:oo 10.00 49 ?O ::1:1: ?92.00 ;; !! 'I:% 12:!lo 8:ot 12.00

106.00 . ...4

t::. 161: 125.00 162. lf4. 129.OC I::.:! 2.50 PAGE “w:” “1’:40°i: coc400~~ 23061: 76lO:OO 9a:oo xi $o:oo

::*:o” 15:oo 45.00 f:% ::3::* 5:oo 23731: 10.00 1130.00

‘xi IS:00

f::l% 99o:oo 1130.00 25;i.g

25.00 119o:oo 1150.00

100.00

::*?i IO:00 15.00

x t::Il.:: 1o:oo 399:oo 1210.00 :!*I: 1460.00 45:oo 26.00 t::x 113:oo 329. 310. 311.

::% ::t+ 20410. 3O:QO 726:Q0 25700.00 15.00 VC OLS 127.00

24.L? 2.50

O.OC

:'*oo" t- 2.50 h*88 02:oo 1g.;g

lb:90

6!.00 tS:*ll 114.00 lio:oo llO.00 f::+ 9s:oo 95.00 Il.00

111.00

1.20

263.00

%+I lr7:oo 16.00 166.00 I-:: 17.00 1:oo ::3 :::+ G:81 Ilo: 194:oo P::8 ““:00”~: 702:QQ

20419. 20221. f%: 0. 26 .MFtiCA!l h 1st llRF 11.00 :::* 16.00 411: :::- :;:: 4161 :::*i: 411. 219:OO 416.

:::. 4.92:

:I:* ::g:

4ta:

:::- 491: -I.