Search

The Bon Accord nickel body has been known since the 1920s to contain rocks with up to 50 wt% NiO. Numerous nickel-rich minerals have been described from this deposit. However, none of these minerals contains significant Cu or S, making the deposit chemically distinct from all other known Ni deposits. The origin of the Bon Accord nickel body is highly contentious, with previous studies suggesting three major possible origins: (1) a hydrothermal origin; (2) an Fe-Ni meteorite that fell into and was altered by an active ultramafic lava flow; or (3) a deep mantle plume that contained a fragment of nickel-rich material. Here, we present new field, petrographic, and geochemical data in an attempt to constrain the origin of this enigmatic body. Based on our fieldwork, there are at least two distinct Ni-rich bodies. Based on the trace-element chemistry, the protolith of the body was a komatiite, likely belonging to the Weltevreden Formation. Because the Ni end member of olivine (liebenbergite) is present in the form of euhedral crystals, this mineral most likely crystallized from a Ni-rich melt. The redistribution of the nickel appears to be due to hydrothermal activity that occurred during the intrusion of the Stentor pluton. Consistent with previous studies, we find that the komatiitic affinity of the host rocks, the stratigraphic controls on the deposit, and the regional distribution of Ni-rich material are inconsistent with a meteorite origin; instead, a komatiite plume sampling a Ni-rich portion of the mantle is currently the best explanation for the origin of the Ni-rich material.

New perspectives on the hydrology, salinization of organic-clay layers, fossilization of faunal remains, and the formation of the Florisbad spring site are presented here. These are based on a critique of previous hypotheses, as well as new evidence and subsequent research. It is contended that the fossilization of faunal remains could not (as was previously postulated) have taken place in areas of spring activity due to the low Ca content of the spring water and a high Eh resulting from spring flow. It is further contended that fossils occurring in areas of spring activity were demineralized. Evidence suggests that fossilization of faunal remains at Florisbad took place in the environs associated with organic-clay layers, which had high Ca concentrations and a low Eh. A new hypothesis is put forward to account for the depositional history of the Florisbad spring site. This accommodates the formation of the site in light of its stratigraphy, geomorphology, salinization processes, and the fossilization of faunal remains. This hypothesis also explains a number of observed geomorphic anomalies.

This paper presents whole-rock geochemical and mineral chemical data for chromitite layers of the Middle Group of the eastern Bushveld complex. It reports on compositional variations of chromite from the chromitite layers, and special attention is paid to the platinum-group element (PGE) content and mineralisation within the chromitite layers. The discussion is focused on possible mechanisms by which the association of PGE enrichment within the chromitite layers can be modelled. The MG chromitites from bottom (MG0) to top (MG4C) are characterised by progressive melt evolution, showing decreasing Mg# and Cr/(Cr+Fe) ratio and increasing Cr/(Cr+Al) ratio, TiO2 and V concentrations. An increase in the Mg# together with a drop in the Cr/Cr+Al) and an increasing Cr/(Cr+Fe) ratio at the base of the MG4A chromitite strongly suggests the addition of hot and primitive magma. High Cu concentrations in the chromitite layers with coincident low S values suggest the application of Naldrett et al. 36 Fe-loss mo...

The geology around the unusual quartz-vein hosted Waterberg platinum deposit was mapped using aerial photographs at a scale of 1 : 1000 to establish possible controls on this unique style of mineralisation. The study area is located in the Mookgophong (Naboomspruit) District of the Limpopo Province in South Africa and is dominated by a gently to moderately, NNW-dipping irregular sequence of bimodal (rhyolite–basalt) volcanics, interbedded with many sedimentary rocks, that form part of the extensive 2˙06 Ga Rooiberg Group, that are cut by variably striking, steep to subvertical quartz veins. The regional geology of the area was mapped at various scales in the 1920s and 1940s, but much of this work remains inaccessible and mapping during this study has uncovered more detail. The major platiniferous quartz vein (known as the 'Main Lode') juxtaposes volcanic rocks of Rooiberg age against Triassic sandstones and therefore occupies a fault. There is no textural evidence that quartz veining was syntecton...

Granulites exposed in the core of the Vredefort Dome are generally thought to represent components of the Archean middle crust of the central Kaapvaal Craton. These rocks were exhumed from depths of at least 20 km following an ~2.02 Ga meteorite impact event, exposing a “crust on edge” section believed to represent a continuous magmatically-differentiated sequence from upper crust possibly down to Moho. Platinum group element concentrations suggest that mafic and ultramafic enclaves in the lower most sections are fragments of greenstone belt material metamorphosed at granulite grade. Re/Os isotopic data are consistent with previously published U-Pb zircon data, which suggest that the mafic and ultramafic rocks have a primary age of ~3.5 Ga and were metamorphosed to granulite grade at ~3.1 Ga. However, near the core of the Vredefort Dome across a major shear zone, other ~3.5 Ga greenstone belt remnants are at distinctly lower metamorphic grade. The juxtaposition of mafic-ultramafic rocks with a different metamorphic history in the core of the Vredefort structure requires a modified “crust on edge” model that incorporates tectonic stacking of different crustal segments at ~3.1 Ga.

We report petrologic, petrographic and chemical data for a newly discovered meteorite specimen from South Africa. The 7.04 kg specimen was found in ca. 1977 on the farm Waterval 64, about 30 km from the town of Merweville (−375 km northeast of Cape Town). Our research group confirmed the identity of the stone as a meteorite in 1997, and obtained a ~1.5 kg portion for study and classification. Externally, the specimen shows a few regmaglypts and a reddish-brown, weathered fusion crust. The sawn internal surfaces show that terrestrial weathering effects, which render a dark brownish colour to the specimen, have penetrated up to 4 to 5 cm into the meteorite. The relatively unweathered interior is light orange-brown in colour. Petrographically, the specimen shows a dominantly recrystallised texture, with a few recognisable chondrules or chondrule fragments. Shock stage is estimated at S2, and weathering grade is W2. Microprobe analyses yield: olivine (Fa23.32 ±0.612, PMD = 0.52, n = 13), orthopyroxene (Fs20.59 ±0.54, n 15), kamacite (Ni = 6.48 ± 0.24 wt.%, Co = 0.94 ± 0.14 wt.%, n = 15), taenite (Ni = 28.9 ± 2.3 wt.%, Co = 0.29 ± 0.08 wt.%, n = 3), troilite (Ni = 0.10 ±0.11 wt.%, n = 17) and chromite (Cr/(Cr+Al) = 0.860 ± 0.003, n = 4). No recognisable feldspar or glass is present in our thin sections of Merweville The dominantly recrystallised but partial preservation of chondrules, apparent absence of glass, and the relative uniformity of olivine and pyroxene compositions are all consistent with classification of Merweville as petrologic type 5. Chemical and mineralogical data for the “unweathered” portion of the meteorite fall best within the L group. We conclude that Merweville is an L5 chondrite. This new specimen increases the inventory of known South African meteorites to a total of 48.

Radiometric dating of melt rocks at impact craters has revealed that some giant impacts appear to overlap in time with major boundaries in Earth history [e.g., the Cretaceous–Tertiary (K/T) and Jurassic–Cretaceous (J/K) boundaries]. The Morokweng impact crater in South Africa is coincident in age with the J/K boundary. However, the types of objects that generate large craters are poorly known because it is difficult to unambiguously identify the projectile from the signature it imparts into the impact rocks. Meteorites are highly enriched in the platinum-group elements (PGE), which have been widely used as a tool for identifying the presence of a meteorite signature. Here we present new PGE analyses from the Morokweng impact melt sheet. Our data reveal high PGE concentrations and high degree of PGE correlation through the melt sheet. Regression analysis was used to determine the projectile PGE signature and constrain input from the terrestrial target rocks. The closest match to Morokweng is the PGE signature of ordinary (L or LL) chondrite meteorites, which is broadly in agreement with the results of an earlier Cr isotope study. The results of these independent studies provide strong evidence that a large, ordinary chondrite projectile struck the area of Morokweng in the late Jurassic.

A magnetic anomaly near the centre of the Vredefort impact structure is attributed to remanent magnetism associated with a large meteorite impact at ~2.0 Ga. The rocks that underlie the anomaly are shocked Archaean granites, which have extremely high remanent magnetism. We propose that the high remanence derives from elongated, micron-size single-domain magnetite particles found along shock-induced planar deformation features in quartz. The magnetite particles appearto have preferred magnetic dipole orientations, and are interpreted to have been intruded alongthe planar deformation features under extreme P—T conditions following the impact event.

A Fe and Mn heterogeneity effect was observed in microscopic garnet inclusion in diamond using micro-PIXE. Results are presented and the implications of these disequilibrium element distributions during growth of garnets are discussed. The occurrence of zoning or any type of chemical inhomogeneity within inclusions encapsulated within diamond has implications on the age of formation of the inclusion and the diamond. It is believed that the age of the inclusion is that of the diamond formation itself, with subsequent isolation of the inclusion from the mantle by the diamond. The significance of the chemical heterogeneity or zoning is that it should not exist given the high temperature and long time of encapsulation.

Most of the statistics given in this overview of the history and current status of Geoscience Education in South Africa pertain to the more geologically inclined disciplines; while the report does mention the extent to which earth sciences are taught in mining and geography departments, no detailed information about these activities are given. There are 13 active geoscience departments countrywide (eleven at universities and two at technical institutions) teaching a wide range of geological topics, some at a highly specialised level. There are just over 100 academic staff members engaged in teaching, supported by 65 technical and administrative staff. Of the teaching staff, 89% have Ph.D. degrees, and most are engaged in active research. About 150 three-year B.Sc., slightly fewer B.Sc. (Hons.), graduates, plus 10 geological technicians pass through the system every year, with most finding employment in the mining industry. Approximately 120 M.Sc. and 60 Ph.D. candidates are currently registered at the universities, about 40% of whom graduate in any particular year.

A roughly 70 km diameter circular feature buried beneath the Kalahari sands in South Africa is revealed on regional aeromagnetic maps. Boreholes drilled into the centre of the structure intercept a ∼ 250 m thick sheet of quartz norite, interpreted as an impact melt, which overlies brecciated and shock metamorphosed basement granite. Zircons recovered from the quartz norite, yield U-Pb ages of 145 ± 0.8 Ma, and biotites provide Ar-Ar ages of 144 ± 4 Ma. These data provide strong evidence for the occurrence of a Late Jurassic impact crater (the Morokweng impact structure) ∼ 100 m beneath the surface.

The determination of ultratrace concentrations of Pt in rock samples using nickel sulphide fire assay combined with neutron activation analysis (NiS-NAA) is influenced by the presence of Te in the irradiated noble metal concentrate. Interference caused by123mTe on the 158 keV peak of199Au (produced from198Pt) is discussed with reference to a noble metal standard (Wits-1) and sedimentary rocks which generally have low Pt concentrations and high Te/Pt ratios. The importance of a rigorously controlled dissolution of the fire assay button to achieve complete removal of Te is emphasized.

Mineralisation in the Merensky Reef of the Bushveld Complex has recently been attributed to the ingress of fluids, possibly via channels beneath depressions in the footwall of the Reef referred to as potholes. In this study samples from immediately below three potholes and from the same horizon (a norite) under normal Reef have been analysed for oxygen and strontium isotopes, chlorine and noble metals to investigate this possibility. Results from one pothole are interpreted as retaining the primary magmatic oxygen isotope signature, with average δ18O of plagioclase and orthopyroxene being +7.1 ± 0.1‰ and +6.6 ± 0.1‰ , respectively, and imply a δ 18 O melt of +6.8‰ Interaction with an early magmatic fluid cannot be detected from the oxygen isotope patterns and therefore the importance of this process cannot be assessed. Results from this oxygen isotope study of the western lobe of the Bushveld Complex are practically identical to those obtained for the laterally equivalent horizons in the eastern lobe. A second pothole shows more variability, with the norite under the pothole Reef showing secondary alteration of plagioclase to paragonite and epidote, and orthopyroxene to serpentine and chlorite. Fresh plagioclase from these altered samples is only slightly depleted in 18O, compared with normal Reef. However, measured whole-rock δ18O-values are up to 1‰ lighter than those predicted from mineral modes, and suggest that the secondary assemblage contains a component depleted in 18O, which could have been the result of exchange with a low-δ18O fluid. This fluid interaction was probably a late-stage, relatively low-temperature event. Secondary alteration of the footwall was accompanied by introduction of chlorine. Au contents also appear to be slightly enriched in the footwall under pothole Reef.

About 30% of the Earth is covered by continents, but only about 10 small kernels of these continentsae—known as Archaean cratonsae—are continental fragments formed before 2.5 Gyr ago. The Kaapvaal craton of South Africa, which formed and stabilized between 3.7 and 2.7 Gyr ago, is one of the oldest reasonably sized examples of these continental fragments. It consists of a mosaic of subdomains that have been welded together by processes similar to those of modern-day plate tectonics. The earliest subdomains may have owed their origin to the onset of efficient recycling from the Earth's hydrosphere into the mantle.

South Africa hosts 70% of the world’s platinum-group element (PGE) resources. Its sustainable management is of utmost economic importance to South Africa and indeed to the world, since the PGE are raw materials for a growing sector of high-tech applications. This holistic project will investigate the chain of natural and industrial processes involved in the formation and utilisation of PGE resources. Scientists from a wide range of disciplines are involved, realizing a first truly multi-disciplinary and multi-facetted scrutiny on the behaviour PGE from ore formation to end-user, and beyond. We intend to use state-of-the-art analytical tools, and theoretical advances, to generate new models for ore formation, which in turn could inform the development of enhanced beneficiation protocols. Furthermore, investigation of the natural leaching and redistribution of the PGE during weathering will be done, and the results used to assess anthropogenic PGE dispersal along highways and relevant industrial sites, with emphasis on their biologic availability. There are three sub-projects, which will focus on investigations into: 1. Primary controls on the natural distribution of the PGE (i) in the mantle, (ii) during magma generation and fractional crystallization, (iii) during magmatic ore-forming processes, and (iv) during post-depositional hydrothermal processes. 2. Sustainable mining, recovery and beneficiation of PGE ores, by quantifying preferred PGE mineralogy and mineral association modes in primary ores, and during the primary flotation circuit. 3. The environmental footprint of PGE recovery and industrial deployment, especially during the weathering cycle. The project extend far beyond the authors and their institutions. Currently, there are six researchers from four institutions in South Africa involved, and four from three German counterparts. Three mining companies have provides samples and/or data and/or support, and three students are already well advanced in their studies.

The Vredefort structure is the oldest and the largest known cryptoexplosion structure on earth. An approximately 36 km deep section through the Archean sialic crust and the overlying Precambrian strata of the Kaapvaal craton is exposed in the core of the structure. The geology presented in the exposed section includes all the principal metamorphic facies in the crust and records a long and complex thermo-tectonic history which dates back to at least 3.5 Ga. The petrographie and geological observations in the basement rocks indicate that there is a complex interrelationship between the Archean geology and the 2.0 Ga dynamic and thermal metamorphic overprint (some of which are postulated to be indicative of impact processes). The dynamic and thermal metamorphic effects do not increase progressively towards the centre of the structure as found at known impact structures. In particular, dynamic deformation phenomena such as pseudotachylite and planar features in quartz reach maximum intensity in the rocks close to the Vredefort discontinuity, a brittle-ductile shear zone which separates upper crustal amphibolite facies rocks from lower crustal granulites. In certain other lithological zones, deformation phenomena are noticeably absent or diminished. We suggest that changes in the physical and chemical properties of the rocks from margin to centre of the basement may account for the variation in the intensity of the 2.0 Ga thermal and dynamic metamorphic effects observed at Vredefort. In conclusion, our overall impression of the Vredefort structure is that it is a relic of an ancient meteorite impact crater, but that there were thermo-tectonic events which occurred both prior to and after the postulated impact event, which complicates the interpretation of its origin.

Isotopic, palaeontological and geochemical details of a drill-core section at Richard's Bay, South Africa, are reported. Foraminiferal evidence shows that the section encompasses a transition from the Maastrichtian to the Danian, one of the very few sites in the Southern Hemisphere on land. The carbon isotope profile of the carbonate fraction shows a large negative excursion at the transition. The corresponding oxygen isotope excursion is largely parallel but less pronounced. It is argued that this Cretaceous-Tertiary (K/T) boundary is characterised by strong and well-identified local influences, seen in the mineralogy, in trace-element data, and in partially parallel carbon and oxygen isotopic signals. A possible scenario is presented in which marine regression is accompanied by increased influence of terrestrial material in this lagoonal environment. This could point towards similar, if often less pronounced, influences in some other well-documented K/T sites.

The central region of the Vredefort structure consists of a semi-circular multi-layered sequence of crystalline rocks which are nearly vertical in attitude, and which increase in metamorphic grade towards the core of the structure. Together with the overlying Precambrian strata, this sequence provides a cross-section through almost the entire crustal section of the Kaapvaal craton (36 km). The upper part of the Vredefort crystalline crust consists of a 3.0-Ga sequence of differentiated felsic rocks in amphibolite facies. The lower crust consists of a complex and heterogeneous (both chemically and isotopically) high-grade metamorphic terrain of charnockites, granulites (mafic and felsic) and supracrustal rocks. The upper crust is separated from the lower crust by the Vredefort discontinuity, a brittle-ductile shear zone characterised by a high concentration of pseudotachylite and brecciated rock. Petrographic, chemical and isotopic evidence suggest that the upper and lower crust have undergone very different styles of evolution. This is indicative of different geological environments prior to their present juxtaposition. We speculate that the upper and the lower parts of the Vredefort crystalline crust were juxtaposed during intracratonic thrusting ∼2.8 Ga ago, and that the entire 36-km section of crust, including the Vredefort discontinuity, were rotated and uplifted into their present vertical position during the 2.0-Ga Vredefort catastrophe.

The international Kaapvaal Craton project was used to provide extensive training and human resource capacity building within the South African geoscience research community. This report discusses the mechanisms of recruitment, success rates and outcomes of the southern African students specifically, as well as the positive aspects and shortcomings of the protocol that was followed. Suggestions are offered to improve future programmes of this nature.

This abstract deals exclusively with problems encountered within an Anglophone learning environment, for the simple reason that this is where the author's personal experience is drawn from. The author would suggest that an analogous from. The author would suggest that an analogous set of problems (and hopefully also an analogous set of solutions!) could be described in the Francophone or Portuguese-speaking countries in Africa.

Ultramafic rocks from the center of the Vredefort structure previously have been interpreted either as Early Proterozoic igneous rocks genetically related to the Bushveld complex, or as Archean upper mantle rocks. Re-Os isotopic data from these rocks display very low 187 Os/ 188 Os and 187 Re/ 188 Os, suggesting that they are not magma chamber cumulates, but rather the residues of melt extraction, similar in Re-Os systematics to subcontinental lithospheric mantle. Whole-rock Re depletion model ages for two samples are 3.3 and 3.5 Ga, indicating that the postulated melt extraction event must have occurred in the Archean. On the basis of these data we conclude that the ultramafic rocks are not related to the Bushveld event, but rather represent upper mantle that rebounded to crustal levels following an impact event. This conclusion strengthens evidence for a model whereby the rocks exposed in the core of the Vredefort structure represent the lower parts of a crust on edge profile that possibly contains a paleo-Moho very near the Earth9s surface.

Diamond inclusions are thought to preserve the characteristics of rocks from beneath the Archean cratons. We have found that refractory trace element ratios in these inclusions are similar to those measured in both rocks of the present-day oceanic lithosphere and some Archean greenstone belts. These observations are consistent with a model in which Archean continental lithosphere evolved by tectonic stacking of oceanic crust and its underlying mantle during intraoceanic collisions. We conceive the continental lithosphere to consist of alternating layers of basalt and peridotite derived from subducted and obducted Archean oceanic lithosphere. This process provides a mechanism for the genesis of diamonds in the metamorphic equivalent of these Archean rocks.

A strong magnetic anomaly near the center of the ancient and deeply eroded Vredefort structure is attributed to remanent magnetization caused by a large meteorite impact at ∼2.0 Ga. The rocks underlying the anomaly are Archean gneisses thought to represent mid-crustal depths that were uplifted to the surface during the postulated impact event. Measurements of the remanent magnetization of the basement rocks yielded consistent vectors of declination = 25° , inclination = 56° , k = 18, α = 16 that correspond to the paleomagnetic pole position at time of impact. Petrologic studies indicate that during impact, large volumes of rock were heated enough to cause thermal remagnetization in the ambient field. Thermal effects of all large impacts on cratons may induce a remanent magnetization of sufficient intensity to cause anomalies in the geomagnetic field that are detectable even by satellites.

Forty whole-rock samples representing the principal silicate rock types of the lower and the lower critical zones of the Bushveld Complex were analyzed for Ir, Ru, Rh, Pt, Pd, and Au. Ir concentration ranges from 0.1 to 2.2 ppb, with sporadic values greater than 4 ppb. There is no readily evident trend of Ir enrichment or depletion up the sequence. Dunites (85-98% olivine) have less Ir (

Previous work has defined three suites of sills and marginal rocks adjacent to and below the Rustenburg Layered Suite of the Bushveld Complex, namely an ultramafic (MgO approximately 30%), a magnesian basaltic (MgO approximately 12%), and a tholeiitic suite (MgO approximately 7%). These three suites are thought to represent magmas which have intruded and been incorporated into the Rustenburg Layered Suite at various times during its formation. The elements Ir, Ru, Rh, Pd, Pt, and Au have been analyzed in 17 samples representative of the three suites. The ultramafic sills have the highest total noble metal content (46 ppb) followed by the magnesian basaltic suite (37 ppb) and the tholeiitic suite (21 ppb). The ultramafic sills are characterized by a relatively fiat chondrite-normalized platinum-group element pattern with an average Pt/Ir ratio of 3.52 and a pronounced Rh anomaly. The magnesian basaltic and tholeiitic suites both have steep positive chondrite-normalized patterns with average Pt/Ir ratios of 16.14 and 40.5, respectively, and no Rh anomaly. Modeling done using incompatible trace elements and petrographic mixing suggests the ultramafic sills represent crystal mushes (of cumulate olivine and chromite) that were intruded into the floor during the crystallization of the lower portion of the Rustenburg Layered Suite. The enrichment of Ir, Ru, and Rh in the ultramafic sills appears to be associated with the cumulate enrichment of olivine or chromite in these sills.The chondrite-normalized patterns found for the Merensky Reef and the Platreef are similar to those of the magnesian basaltic magma. The pattern found for the UG-2 is much flatter and in some respects very similar to that found for the ultramafic sills.On a worldwide basis the platinum-group element contents of the magnesian basaltic and tholeiitic magmas generally fall in the range of basalts and chill zones of other layered suites. The magnesian basaltic magma does have one of the highest Pt contents measured for a magma.

Frequently samples are presented in powdered form, or have for reasons of inhomogeneity to be powdered and thoroughly mixed in order to get a representative sample. In other situations while solid samples may be available, in the case of precious minerals such as diamond, they may be very small so that accurate results can inherently not be obtained. Since these problems frequently present themselves with our interests in nuclear geology, solutions to the problem which does not arise in most other analytical techniques have been sought and now established. Two basic approaches will be described together with the instrumentation developed.

The elevated concentration of iridium-one of the platinum-group elements (PGE)-at the Cretaceous-Tertiary boundary is still the most generally accepted evidence that a large bolide struck the earth at the time of the end-Cretaceous mass extinctions. New chemostratigraphic data for cross-boundary sections from both hemispheres are not easily explained in terms of such an impact event, for example the observation that the PGE patterns show marked differences between the hemispheres. The new constraints indicate that models of mantle-derived PGE should be seriously considered, and that PGE anomalies might not be as useful as previously thought as unambiguous identifiers of large impact events in the earth's history.

The development of ion beam techniques, namely particle induced X-ray emission (PIXE) and charged particle activation analysis (CPAA), for the determination of the platinum group elements (PGEs) in terrestrial rocks is described. The noble metals are heterogeneously distributed in rocks and ores which necessitates that large sample sizes be analysed to obtain representative results. A nickel sulphide fire-assay technique has been utilized to concentrate the noble metals from a large sample (50 g). The concentrated noble metals are finally retained on a filter membrane that is used as the target for ion beam analyses. Limits of determination of 0.5 ng/g were established for ruthenium, rhodium and palladium and 1 ng/g for platinum by PIXE; whereas osmium could be determined at 0.5 ng/g levels by CPAA. The ion beam techniques compare most favourably in terms of rapidity, expense and sensitivities for the platinum group elements, with other analytical techniques.

A failed continental rift system, consisting of several partly discontinuous grabens, developed during Middle Proterozoic times in SWA/Namibia and Botswana, possibly in response to a mantle plume. The basins contain thick successions of bimodal volcanics, and continental and marine sediments. High-potassium acid volcanics developed as a result of partial melting of mainly sedimentary continental crust. The basal acid volcanics are enriched in PGE and Au which might have been supplied by mantle derived fluids, released during the initial phase of crustal melting and fracturing of the lithosphere. After this anomalous phase of acid volcanism, ‘more normal’ acid volcanics were extruded. Mantle derived tholeiitic basalts, which were extruded along deep reaching fractures, show typically magmatic PGE and Au distribution patterns. Overlying coarse continental red beds, derived from the denudation of the older basement and the reworking of uplifted and blockfaulted acid volcanics, are generally barren of PGE and Au. Laterally extensive, sediment-hosted, stratabound Cu/Ag mineralization at the base of a dark pyritic shale unit is slightly enriched in Pt (0·04–0·12 g/t), Au (0·04–0·06 g/t), Th and Ni. Generally, the bimodal volcanics display typical tholeiitic chondrite normalised PGE patterns. Altered volcanic rocks, as well as the ore samples, show PGE patterns which may indicate hydrothermal reworking. The authors suggest a model wherein the PGE and Au are mantle derived, possibly from the mantle plume which triggered the development of the continental rift system. Selective remobilisation leached Cu, Zn and Co from the basalts and Pt and Au from the red beds. The metals precipitated at the redox-cline between the pyritic sediments and the red beds, whereas overlying sediments, also pyritic, remained unmineralised.

PGE and other siderophile anomalies at the Cretaceous-Tertiary (K-T) boundary have been interpreted as evidence for catastrophic extraterrestrial influx. We list some key points from the literature which are difficult to resolve in terms of this model: (i) The relative concentrations of siderophile elements cannot be explained in terms of any known meteorite composition; (ii) The concentration of Re, Ba, As, Sb, Mo, Zn and Cd is high, relative to chrondrite (and usually interpreted to be of terrestrial origin) and Cr, As, Sb and the REE are high relative to iron meteorites; (iii) Stable isotopes of the noble gases suggest a terrestrial source; (iv) osmium isotopic studies cannot distinguish between a meteoritic or mantle component.

WITS-1 is a sample of silicified komatiite collected from close to the Komati River type section, near Barberton in South Africa. Two thousand kg of this komatiite was crushed and prepared for standardization. Noble metal analysis of WITS-1 has been carried out by six laboratories using four different analytical methods. The data obtained from these sources were subjected to statistical analysis (F ratio and Student t test) and a suite of “preferred” values for the noble metal concentrations are reported.

Three major sulfide zones containing platinum group elements (PGE) have been identified in the upper part of the ultramafic sequence of the Great Dyke. The PGE are at the bases of the sulfide zones; the lower two have sparse sulfide mineralization over a thick interval; the upper zone contains higher grade in a thinner stratigraphic layer. None of the current models for the formation of stratabound sulfide deposits in layered intrusions adequately accounts for the mineralization. The two lower sulfide zones are interpreted as resulting from progressive fractional segregation of sulfide from the silicate magma, whereas the upper zone may have resulted from mixing of a relatively primitive magma at the base of the chamber with an evolved liquid derived from near the roof.

The 3·5 Ga mafic-ultramafic rocks of the Barberton greenstone belt, South Africa, form a pseudostratigraphy comparable to that of Phanerozoic ophiolites. This Archaean complex, referred to here as the Jamestown Ophiolite Complex, consists of high temperature tectonometamorphic peridotites overlain by an intrusive-extrusive igneous section, which in turn is capped by a chert-shale sequence. Elevated PGE concentrations with distinctly different chondrite normalized patterns have been found at three pseudostratigraphic levels of this ophiolite. From the top down, high PGE values are first encountered in an undeformed chromitite, interlayered with meta-pyroxenites. The PGE abundances and normalized patterns are similar to those reported from PGE mineralized zones in large mafic layered complexes. Chromitemagnetite pods occur stratigraphically below the chromitite. Their PGE patterns resemble those reported from chromite pods in Phanerozoic ophiolites and those of the overlying chromitite. The host rocks of the pods (deformed talc-carbonate rocks) are interpreted as a mixture of deformed high-temperature cumulates and depleted upper mantle residue, within a ductile shear zone. In all the above occurrences, the PGE were concentrated by igneous upper mantle-crustal fractionation processes of mantle derived liquids. The third type of PGE concentration is found in an unusual, sulphur-poor NiFe-silicate-spinel pod (called the Bon Accord-[BA]-deposit) and an associated hematite-magnetite-chromite (HMC) pod, both of which occur in coarse-grained meta-dunites at the lowest levels in the ophiolite. These host rocks are interpreted as high-temperature tectonites (strongly depleted in REE); they probably represent depleted upper mantle residue from which the overlying rocks were derived by partial melting. BA can be divided into two groups. Group A has pat, chondritelike PGE patterns. Group B has a lower concentration of all the PGE, but the relative depletion is most severe for Os and Ir (which are the most refractory); thus the patterns have a positive slope. Both BA groups have negative Pt anomalies in their PGE patterns (more pronounced for Group A), while the HMC-pod, which is depleted in the other PGE, relative to BA, has a strong positive Pt anomaly. It is difficult to explain these fractionations in terms of known upper mantle-crustal processes, especially in the absence of sulphur. The PGE differentiation of BA is interpreted as an indication of chemical fractionation which started in the lower mantle. We conclude that the body represents a metal-silicate sample derived from an almost chondrite-like lower mantle where it resided as a left-over from inefficient core formation. This siderophile-enriched mantle fractionated during ascent and some of its residue was incorporated into the lower portion of the lithosphere during formation of the Archaean oceanic crust. Tectonic and chemical data suggest that the thick and old ‘lithospheric keel’ beneath the South African craton has a large component of Archaean residual sima and is therefore likely to be relatively enriched in PGE.

Closely spaced (cm-scale) traverses through the K-T boundary at Stevns Klint (Denmark), Woodside Creek (New Zealand) and a new Southern Hemisphere site at Richards Bay (South Africa) were subjected to trace element and isotopic (C, O, Sr) investigation. Intercomparison between these data-sets, and correlation with the broad K-T database available in the literature, indicate that the chemistry of the boundary clays is not globally constant. Variations are more common than similarities, both of absolute concentrations, and interelement ratios. For example, the chondrite normalized platinum-group elements (PGE) patterns of Stevns Klint are not like those of Woodside Creek, with the Pt/Os ratios showing the biggest variation. These differences in PGE patterns are difficult to explain by secondary alteration of a layer that was originally chemically homogeneous, especially for elements of such dubious crustal mobility as Os and Ir. The data also show that enhanced PGE concentrations, with similar trends to those of the boundary layers, occur in the Cretaceous sediments below the actual boundary at Stevns Klint and all three the New Zealand localities. This confirms the observations of others that the geochemistry of the boundary layers apparently does not record a unique component. It is suggested that terrestrial processes, eg. an extended period of Late Cretaceous volcanism can offer a satisfactory explanation for the features of the K-T geochemical anomaly. Such models would probably be more consistent with the observed stepwise, or gradual, palaeontological changes across this boundary, than the instant catastrophe predicated by the impact theory.