no abstract available

 An unusual association of platinum-group minerals (PGM) is observed in a chromitite layer of the Mount Bol' shaya Varaka deposit, Imandra layered complex, Kola Peninsula, Russia. The PGM typically occur as cluster-like. or chain-like micro-aggreates. and as separate minute g-rams located in contact with chromite along grain boundaries. The PGM are enclosed in magnesiohornblende, edenite, talc, or phlogopite, and may transect the host hydrous silicates; they include sperrylite, laurite-erlichmanite. unnamed Cu-2(Ni,Co)Pt3S8. hollingworthite, pyrite rich in the platinum-group elements (Ru 10.96, Os 0.98, and Ir 0.54 wt.%), nickelian platarsite,(Pt0.71Ni0.26Co0.05)(Sigma1.02)As1.10S0.88, daomanite, and cooperite. Concentrations of Rh and S in the rhodian sulfurian sperrylite vary front 1.0 and 0.8 to 6.6 and 4.9 wt.%, respectively, and the composition most enriched in Rh and S is (Pt0.78Rh0.20Ir0.03)(Sigma1.01)(As1.53S0.47)(Sigma2.00). The correlations Pt-Rh (correlation coefficient R = -0.99), Pt-As (R = 0.98), PtS (R = -0.99). Rh-As (R = -0.97), Rh-S (R = 0.97), and As-S (R = -1.00) are important and consistent with the existence of a new series of solid solution, extending from PtAs2 toward the pyrite-type Rh1-xS2 ("Rh2S5"). Most of the PGM precipitated during deuteric alteration at a postmagmatic-hydrothermal stage of crystallization of the chromitite. The majority of the laurite-erlichmanite grains display a close textural relationship with the hydrous silicates and form part of the hydrothermal mineralization at Imandra.

 Electron-microprobe analyses of the dunite and olivine chromitite of the ultramafic complex at Goodnews Bay, Alalska, show that the chromite contains more ferrous and ferric ion than is found in chromite from Alpine or stratiform complexes. The iron-magnesium distribution coeficients between chromite and olivine demonstrate that more ferrous iron is partitioned to the chromite in the Goodnews Bay comples than to the chromite in stratiform chromitite but that the distribution is similar to that for Alpine chromitite. The high iron content of the chromite and olivine at Goodnews Bay and the general geology of the complex indicate that it is a concentric-type ultramafic complex similar to the platinum-bearing concentric complexes of southeastern Alaska. A platinum-alloy inclusion in the chromite found in the ultramafic section at Goodnews Bay demonstrates that the olivine chromitite is the source of at least some of the platinum.

 An expansion of alpine glaciers during the latest Pleistocene produced an extraordinarily well defined end moraine system in the Ahklun Mountains, southwestern Alaska. These moraines, deposited during the Mount Waskey advance, are several kilometers beyond modern glacier termini, and similar to80 km upvalley of the late Wisconsin Ahklun Mountains ice cap terminal moraine. Eleven cosmogenic Be-10 and Al-26 exposure ages on moraine boulders, combined with radiocarbon ages from a lake core upvalley of a moraine deposited during the Mount Waskey advance, suggest that the advance culminated between 12.4 and 11.0 ka, sometime during, or shortly following, the Younger Dryas event (ca. 12.9-11.6 ka). We believe that the Mount Waskey advance was a consequence of cooling during the Younger Dryas. These data further strengthen emerging evidence for Younger Dryas-age cooling of the North Pacific region.

 A multilayer feed-forward neural network, trained with a gradient descent, back-propagation algorithm, is used to estimate the favourability for gold deposits using a raster GIS database for the Tenterfield 1:100 000 sheet area, New South Wales. The database consists of solid geology, regional faults, airborne magnetic and gamma-ray survey data (U, Th, K and total count channels), and 63 deposit and occurrence locations. Input to the neural network consists of feature vectors formed by combining the values from co-registered grid cells in each GIS thematic layer. The network was trained using binary target values to indicate the presence or absence of deposits. Although the neural network was trained as a binary classifier, output values for the trained network are in the range [0.1, 0.9] and are interpreted to indicate the degree of similarity of each input vector to a composite of all the deposit vectors used in training. These values are rescaled to produce a multi-class prospectivity map. To validate and assess the effectiveness of the neural-network method, mineral-prospectivity maps are also prepared using the empirical weights of evidence and the conceptual fuzzy-logic methods. The neural-network method produces a geologically plausible mineral-prospectivity map similar, but superior, to the fuzzy logic and weights of evidence maps. The results of this study indicate that the use of neural networks for the integration of large multisource datasets used in regional mineral exploration, and for prediction of mineral prospectivity, offers several advantages over existing methods. These include the ability of neural networks to: (i) respond to critical combinations of parameters rather than increase the estimated prospectivity in response to each individual favourable parameter: (ii) combine datasets without the loss of information inherent in existing methods: and (iii) produce results that are relatively unaffected by redundant data, spurious data and data containing multiple populations. Statistical measures of map quality indicate that the neural-network method performs as well as, or better than, existing methods while using approximately one-third less data than the weights of evidence method.

 Mineralogical studies of platinum-group minerals (PGM) made during the period 1970-1995 are compiled in this paper for PGM obtained from platinum- and gold-bearing placer deposits of fifteen countries. A total of 3399 quantitative electron microprobe analyses of 37 different PGM are presented, together with analytical data plotted on compositional diagrams. The data reflect the dominance of Pt-Fe alloys (1723 analyses) and Os-Ir-Ru-Pt alloys (976 analyses) in the placers. The platinum-group element (PGE) alloys, as well as some other PGM, generally occur as grains less than one millimetre in size. The Pt-Fe alloys, containing some Cu and Ni, are characterized by a wide range of Pt:(Fe, Cu, Ni) ratios and, where sufficient data are available, often show a bimodal population at about 16-17 and 25 at.% (Fe, Cu, Ni). Iridium, osmium, and rhodium occur in trace quantities in over 96% of the analyzed grains, and all three elements were never found to be in concentrations below their detection levels simultaneously. The large data base for the hexagonal alloys shows a well-developed trend parallelling the miscibility gap in the Os-Ir-Ru ternary system. New analyses have narrowed further the width of this miscibility gap, which we postulate to be due to the formation of some alloys at higher temperatures. Important features of the PGM alloys studied include the presence of inclusions of other PGM (as well as some undefined PGM), PGE-bearing minerals, spinels, silicates, and more rarely sulfides such as chalcopyrite, bornite, and pentlandite, as well as their crystallographic orientation and textures. Alteration and replacement features (when present) are minor, not pervasive, and usually restricted to a limited number of PGM. The sum of the features obtained by detailed mineralogical characterization of the placer PGM is consistent with a high temperature intrusive origin. Evidence of minor secondary alteration that is sometimes found (e.g., irarsite, tulameenite) is probably associated with later processes such as serpentinization. The presence of PGE-oxides and hydroxides,all of which are still incompletely characterized, is ascribed to surficial weathering processes. In many cases the placer PGM can be traced to the source intrusive rocks (e.g., zoned ultramafic complex, ophiolite). However, correlation with source rocks is sometimes not possible due to deep weathering, lack of detailed geological maps, or inacessible field areas. Copyright (C) 1996 Canadian Institute of Mining, Metallurgy and Petroleum.

 Remote sensing techniques and spatial data analysis through Geographic Information Systems (GIS) have been jointly applied in a mineral exploration context to identify gold-rich potential areas in SE Spain. Results confirm the usefulness of this integrated methodological approach as an effective tool to assess mineral potential in the studied region. Satellite and airborne image analysis have offered valuable thematic information referring both to lithology and altered zone mapping from photointerpretation and digital classification. For this goal, SPOT panchromatic and Landsat TM images were merged in order to obtain high resolution image documents for photointerpretation purposes, and the Feature Principal Component Selection technique was applied to highlight hydrothermal alteration zones characterized by hydroxyl-bearing minerals. Remote sensing results were integrated, in conjunction with existing maps and data from mineral exploration surveys, into the GIS as vector or raster layers. The GIS implementation stage consisted of the creation of a relational database, including a complete set of georeferenced data, and the elaboration of a specific user interface for spatial analysis using GIS facilities in order to establish a Decision Support System (DSS). The system was used to simulate different mineral exploration scenarios, calculating and mapping a Mineral Potential Index that was interpreted in a practical sense in terms of surface reduction and economic costs.

 There are 96 approved PG (platinum group) - mineral species and more than 500 unidentified PG-phases on record. In addition, over 20 non-PG-minerals contain varying concentrations of one or more of the six platinum group elements. The PG-minerals commonly have cubic symmetry and density Values mostly between 10-11 g/cm(3) but within the range 5-22 g/cm(3). Furthermore, there is a tangible positive correlation between relative density and reflectance data. On a chemical basis, all PG-minerals and PG-phases are non-silicates and a large proportion are sulphides, arsenides and tellurides. The elements Sb and Bi are also important constituents of PG-minerals. This review of the published literature up to April, 1996, provides an evaluation of approximately 1500 published chemical analyses of PG-minerals which have been accepted by the IMA as discrete mineral species and unidentified PG-phases reported in the literature but not proven to be discrete minerals. Revealed by the database are a number of chemically-related trends which concern both the PG-elements themselves and the essential non-PG-element constituents. The latter grouping of elements is discussed as a separate publication (Daltry and Wilson, 1997). Pd-dominant compounds are the most abundant of the PG-minerals and PG-phases and the Os-dominant compounds are the least abundant. On an inter-PG-element basis, the Ir-, Os-, Pd-, Pt-, Rh-and Ru-dominant minerals and phases may each contain varying concentrations of all the remaining five PG-elements. In this context, there are discernable sympathetic and antipathetic trends. Palladium exhibits the most pronounced antipathy with other PG-elements, whereas platinum exhibits the most marked sympathetic tendencies towards the other PG-elements.

 no abstract available

 The Quaternary gold-platinum-group mineral (PGM) placers of the Aluchin horst, western Chukotka, Russian Far East, contain PGM (mostly Pt-Fe alloys) typical of those associated with Alaskan-Ural-type complexes. Such complexes intruded Mesozoic sedimentary rocks. They form the primary source of PGM. The PGM placer nuggets vary considerably in size (0.1-9 mm), shape, and degree of abrasion, indicating a long history of placer-forming processes. Some of the nuggets are covered by shells of detrital minerals (mostly Al-Si phases) indicating derivation from previous host sediments of the Volga (Late Jurassic) stage. The quartz inclusions observed in Pt-Fe alloys are detrital and have been pressed into the nuggets during lithification of the Volga sediments. Concentration of the PGM took place in two stages: during erosion of the Alaskan-Ural-type intrusions and deposition of PGM-bearing detritus in the lowermost Volga Series; and recycling of the Volga sediments and deposition of the PGM in placers during the Quaternary.

 The nomenclature of Os-Ir-Ru alloys, as revised by its Nomenclature Subcommittee, has been approved by the Commission of New Minerals and Mineral Names, International Mineralogical Association. Henceforth, only four names apply to minerals whose compositions (in at.%) lie within the ternary system Os-Ir-Ru: osmium for all hexagonal alloys with Os the major element; iridium for all cubic alloys with iridium the major element; rutheniridosmine for all hexagonal alloys with Ir the major element; and ruthenium for all hexagonal alloys with Ru the major element. The nomenclature of alloys of platinum-group elements in the ternary systems Os-Ir-Pt, Ru-Ir-Pt, Ir-Ru-Rh, Ir-Os-Rh and Pd-Ir-Pt are reviewed.

 The spatial distribution of particulate gold in sediments blanketing the Nome nearshore area, Alaska, is the product of diverse late Cenozoic glacial, marine, fluvial, and deformational processes. Analysis of the extent of amino acid diagenesis in fossil molluscan shells from 51 boreholes drilled in 1967 by the U.S. Department of Interior in the Nome nearshore area demonstrates that the upper 20 m of the offshore deposits have been pervasively reworked. Below 20 m, sediments accumulated more continuously in a marine environment and contain minor amounts of particulate gold. The initiation of extensive reworking approximately coincides with the Beringian marine transgression (approximately 3.0 my), after which time glaciers advanced southward beyond the present-day coast and left deposits associated with the richest concentrations of particulate gold. Within the area extending approximately 5 km seaward of Nome, gold values in the top 20 m of sediment do not change seaward. The greatest concentrations are found within 4 m of the seafloor; lag-gravel deposits formed on till of the present seafloor are especially enriched. Although erosional surfaces analogous to the present seafloor may be buried within the sequence of Pliocene and Pleistocene deposits, the extent of the gold-bearing glacial deposits that would support these surfaces is not yet fully delineated. Reconstructions must take into account the possibility that glaciers advanced onto the coastal plain during a period of high relative sea level. Coarse-grained, potentially gold-bearing debris would have been dispersed farther beyond a glacier terminating in a marine environment than from terrestrial-based ice. A better understanding of the geologic processes that formed the sedimentary sequence on- and offshore of Nome and consideration of theoretical constraints of placer-gold formation will lead to more reliable predictions of the distribution of particulate gold concentrations.

 Pleistocene glacial-estuarine sediment deposited in an intertidal environment of northeastern Bristol Bay, southwestern Alaska, was dated using a variety of approaches, including infrared stimulated and thermoluminescence (IRSL and TL) techniques. Analysis of modern and C-14-dated Holocene tide-pat mud demonstrates that the bulk of sediment in this environment is reset by solar radiation, thereby lending confidence to ages obtained from similar Pleistocene deposits by luminescence techniques. IRSL seems to be especially well suited for dating, with resolution on time scales of < 10,000 yr. The ages of tide-flat mud of the Nushagak Formation, derived from the Ahklun Mountains to the northwest of Bristol Bay, and of Halfmoon Bay drift, derived from the Alaska Peninsula to the southeast, suggest contemporaneous glacial-estuarine deposition related to independent glacial source areas about 75,000-80,000 yr ago. This age is consistent with other geochronological data that indicate a pre-late-Wisconsin and post-substage-5e age, including nonfinite C-14 ages, a lack of interglacial indicators, and Old Crow tephra (similar to 140,000 yr) atop the drift, normal paleomagnetic inclinations, and amino acid (isoleucine) epimerization ratios (aIle/Ile). AIle/Ile ratios in Portlandia arctica (0.052 +/- 0.003) from a marine-lag horizon at South Naknek beach, which separates Halfmoon Bay drift above from older glacial-estuarine drift below, are only slightly higher than in Mya truncata (0.041 +/- 0.007) from last-interglacial Pelukian deposits at Nome. As laboratory heating experiments show that the two genera epimerize at similar rates, these data imply correlation of the marine lag at South Naknek beach with Pelukian deposits. Hence, glaciers on the Alaska Peninsula experienced major pre-late-Wisconsin advances both before and after the last interglaciation. Shells reworked into Halfmoon Bay drift yield aIle/Ile ratios of 0.028 +/- 0.005 for Portlandia at Second Point and 0.027 +/- 0.001 for Hiatella arctica at Etolin Point. Together with assumptions about the postdepositional temperature history, these ratios indicate that the shells are at least 55,000 yr, and probably closer to similar to 90,000 yr, although the uncertainty in this age estimate is broad. The amino acid and luminescence data converge on an age between about 75,000, and 90,000 yr, late during oxygen-isotope stage 5, for a major ice advance far beyond late-Wisconsin limits. (C) 1996 University of Washington.

 New stratigraphic and geochronologic data from the Togiak Bay area of southwestern Alaska indicate that glaciers advanced from the southern Ahklun Mountains at least three and as many as six times prior to the late Wisconsin. The oldest glaciations are represented by glacial-marine sediment in coastal exposures on Hagemeister Island. The extent of amino acid (isoleucine) epimerization in fossil molluscs indicates that at least one, and possibly four, older middle Pleistocene glacial intervals are represented, with age estimates spanning similar to 500-280 ka and averaging similar to 400 +/- 100 ka. The youngest glacial-marine drift on Hagemeister Island may correlate with the eruption of the Togiak tuya. A new Ar-40/Ar-39 age on basalt that overlies pillow lava indicates that the volcano erupted through glacial ice at least 300 m thick 263 +/- 22 ka. The youngest drift in the region overlies the Old Crow tephra (140 +/- 10 ka) and a 70 +/- 10 ka basaltic lava flow dated by thermoluminescence analysis of underlying baked sediment. The drift delimits flat piedmont lobes that spread out onto the continental shelf and terminated > 100 km from their source areas during the early Wisconsin (sensu late). The glacial-geologic evidence suggests that major expansions of glaciers were out of phase with global ice volume. (C) 2000 Elsevier Science Ltd. All rights reserved.

 An 18-m-high coastal bluff at Togiak Bay (northwestern Bristol Bay, southwestern Alaska) exposes marine, lacustrine, fluvial, glacial, volcanic, and organic deposits that record the similar to 50,000-year-long transition from the peak of the last interglaciation to the early Wisconsin glaciation. The base of the section is dominated by stratified sand and silt extending up to 4.3 m above sea level: marine diatoms are present, and pollen assemblages are characterized by relatively high percentages of Picea, Alnus, and Betula and low percentages of Poaceae and Cyperaceae. The marine sediment was probably deposited during the peak of marine oxygen-isotope stage (OIS) 5e. An infrared stimulated luminescence (IRSL) age of 151,000 +/- 13,000 yr from near the base of the exposure is permissive of this correlation. The marine sand and silt are overlain by 0.8 m of peaty silt with diatoms that record a transition from marine to lacustrine conditions. During this interval, Poaceae and Cyperaceae dominate the pollen assemblages, and Picea and shrubs are nearly absent, suggesting that herb tundra occupied the landscape. This interval probably encompasses OIS 5d on the basis of the herb tundra and an IRSL age of 119,000 +/- 10,000 yr from 60 cm below the marine/lacustrine transition. The organic mud is overlain by 3.1 m of stratified sand and organic silt that apparently record shallowing of the lake; reappearance of spruce and shrubs (=OIS 5c?); and subsequent deepening of the lake (=OIS 5b?); followed by aggradation of a floodplain (=OIS 5a?), which was dry at the time basaltic lava buried the site. Thermoluminescence analyses on lava-baked sediment indicate that the eruption occurred 70,000 +/- 10,000 yr ago. Sometime thereafter, but prior to 53,600 C-14 yr B.P., an outlet of the Ahklun Mountains ice cap advanced over the site and deposited similar to7 m of bouldery ice-contact drift. The sedimentary sequence contains at least four tephra beds. Major- and trace-element chemistry provide a basis for correlating two of the tephras with tephra beds at nearby sites. The tephras, luminescence ages, and correlations with marine isotope stages provide the geochronological control to place the Togiak Bay section into a global context. The site serves as an important new reference section for late Pleistocene paleoenvironmental change in eastern Beringia. (C) 2001 University of Washington.

 The mode of material transportation and differentiation in the coastal zone of shelf is analyzed with the Val'kumei coastal-marine tin placer (Chaun Gulf, East Siberian Sea) as example. The placer incorporates two zones of cassiterite concentration. The first zone with coarse-grained cassiterite is located within the beach and adjacent coastal water area. The second zone with relatively fine-grained cassiterite is located on the submarine coastal slope in the distal area. This type of placer structure reflects the polymodal grain-size distribution of cassiterite in the bedrock source and peculiarities of material distribution in the alongshore debris flow. The proposed mathematical diffusive-convective model of coastal-marine placer formation adequately reflects the real placer structure and makes it possible to predict placer parameters in unilateral (alongshore) debris flow areas based on the composition of placer-forming source and lithodynamic parameters of the coastal water area.

 The performance of five published methods for extracting relative spectral emissivity information from thermal infrared multispectral data has been evaluated. In the first part of this article, we have recalled those five methods and shown mathematically that they are almost equivalent to each other. Then, using simulated data built up wit the TIMS (Thermal Infrared Multispectral Scanner) instrument, we have analyzed the sensitivity of those methods to different sources of error which may occur in real data such as errors due to 1) method simplification, 2) instrumental noise and systematic calibration error, 3) uncertainties on the estimation of downwelling atmospheric radiance, and 4) uncertainties of atmospheric parameters in atmospheric corrections. In terms of resulting errors in relative emissivity, the results show that: a) All methods are very sensitive to the uncertainties of atmosphere. An error of 20% of water vapor in midlatitude summer atmosphere (2.9 cm) may lead to an error of 0.03 (rms) for Channel 1 of TIMS. b) The effect of the atmospheric reflection term is very important. If this term is neglected in method development, this may lead to an error of 0.03 (rms) for Channel 1 and midlatitude summer atmosphere. This is the cae for the alpha method. c) Instrumental noise commonly expressed by noise equivalent difference temperature (NE Delta T) from 0.1 K to 0.3 K results in an error of relative emissivity ranging from 0.002 to 0.005 for all methods. d) Error in relative emissivity due to the instrument calibration error (systematic error) is negligible. The stud also shows that the relative emissivity derived wit an incorrect atmosphere is linearly related to its actual value derived with correct atmospheric parameters. Based on this property, we propose three methods to correct for the errors caused by atmospheric corrections under horizontally invariant atmospheric conditions. A practical analysis with the real TIMS data acquired for Hapex-Sahel experiment in 1992 supports the results of this simulation. (C) Elsevier Science Inc., 1999.

 Four new AMS C-14 ages, glacial mapping, and measurements of soil development, loess thickness, and moraine morphology constrain the age and extent of at least three Quaternary advances by outlet lobes of an ice cap over the Ahklun Mountains of southwestern Alaska. The relative-age data are from 107 sites correlated to 25 ice-marginal positions in the Kanektok, Goodnews, Togiak, and Kulukak River valleys and along the southeastern flank of the range. Radiocarbon ages provide minimum ages for six - and a maximum age for one - of the former ice margins. Soil and morphometric parameters subdivide the ice limits into three relative-age groups. One to three pre-Wisconsin advances, probably middle Pleistocene in age, are represented by drift with relatively thick B horizons (60 +/- 5 cm, with Bt horizons), thick loess caps (80 +/- 12 cm), and broad moraines (135 +/- 134 m) with gentle side slopes (7 +/- 5 degrees). An extensive early Wisconsin (sensu late, s.l.) advance, > 39.9 ka, and three associated stillstands or readvances are characterized by intermediate soil and morphometric parameters (Bw and weak Bt horizons with thicknesses of 40 +/- 11 cm; loess thicknesses of 69 +/- 36 cm; crest widths of 38 +/- 13 m; and slope angles of 14 +/- 4 degrees). The maximum late Wisconsin advance, > 16.9 ka, and two readvances or stillstands are associated with thin (20 +/- 5 cm), weakly to moderately developed Bw horizons, thin loess caps (28 +/- 8 cm), narrow crest widths (28 +/- 9 m), and steep slope angles (18 +/- 3 degrees). The data confirm that early-Wisconsin glaciers in southeast Beringia were much more extensive than late Wisconsin glaciers, which were apparently limited by availability of moisture. (C) 2000 Elsevier Science Ltd. All rights reserved.

 Platinum-group elements (PGE) are typically associated with mafic and ultramafic intrusive rocks and the main exploration targets are layers and zones rich in POP-bearing sulphides. Some PGE occurrences, however, are in sulphide-poor situations and this raises the possibility that PGE deposits may be present in parts of mafic and ultramafic intrusives currently considered to have low exploration potential. A multidisciplinary study was undertaken on four subeconomic deposits of platinum-group metals to develop a model of formation for PGE deposits lacking significant base-metal sulphides. Two of the deposits occur in Albania, in the Tropoja and Bulqiza massifs, and are part of an ophiolitic belt created in an oceanic environment during the Upper Jurassic. The other two deposits occur in Madagascar, in the Andohankiranomena and Lavatrafo ultramafic massifs, and are within a Pan-African rifted zone. A Pt-rich chromitite style of mineralization was identified in the Andohankiranomena and Tropoja deposits, where the PGE are mostly included in chromite. A Pt- and Pd-rich silicate (dunite) style of mineralization was identified in the Lavatrafo and Bulqiza massifs, where PGE mineralization is associated with interstitial material between olivine grains. The four deposits have contrasting patterns of PGE distribution and individual element ratios, suggesting that different mineral species (alloys, arsenides and sulphides) host the PGE. No primary geochemical halos were detected around any of the deposits and weathering has little effect on the distribution of the PGE. The study showed that alloys and arsenides are the main carriers for platinum in all the deposits. Pt-Fe alloys, in particular, are often present in PGP deposits poor in base-metal sulphides and two phase systems were investigated experimentally: Pt-Ir-Fe-S (and the related subsystem Ir-Fe-S) and Fe(Cu)-Pt-Rh-S. A comparison of experimental results with natural phases in the deposits suggests that fluid-assisted exsolution of Pt, Ir and other elements from original higher-temperature solid solutions could be widespread. This supports the fluid-driven multistage mineralization concept suggested by field data. The experimental work also indicated that Pt-Rh-Fe alloys can coexist with two types of immiscible sulphide melt (one low in Rh and high in Cu, the other Rh-rich); both types of sulphide melt may have been present in the Tropoja deposit, where Rh-rich minerals occur. Within all the deposits locally high sulphur fugacity conditions may have developed even in low-sulphur assemblages during the final stages of PGE deposition. In-situ sulphur isotope data from both the Madagascar and Albanian intrusions are compatible with a mantle origin, as are oxygen isotope data on silicates and oxides. Radiogenic isotope data indicate strongly the influence of a crustal component on the Madagascar intrusions. The existence of two distinct radiogenic components at the time of the Albanian PGE mineralization is suggested by some Re-Os isotopes, and this is best interpreted as mantle-derived heterogeneity in the melt from which they formed. The mineralization was subjected to retrograde hydrothermal metamorphism and recrystallization during the cooling of the intrusions. in Madagascar re-equilibration was first to granulite-facies and later to greenschist-facies conditions. A chloride-rich mineralogy resulted from greenschist hydrothermal metamorphism and the fluid movement could explain the absence of sulphides (possibly attacked by a secondary C1-rich fluid), the substantial change of the metal/S ratio and local PGE remobilization. In all the deposits, whether chromitite-style or silicate-style, the PGP metals are concentrated at distinct horizons within the basic-ultrabasic complexes. The field and laboratory studies suggest that both styles of mineralization were produced by a fluid-driven, multistage process within a fractionating silicate magma chamber. Crystallization of Pt-barren massive chromitite was followed by crystallization of Pt-rich chromitite and, finally, by deposition of Pt- and Pt-rich silicate rocks associated locally with gold and disseminated sulphides. The later stages of Pt and Au concentration were due to metal-rich fluids issuing from an ascending intercumulus melt. Once formed, the mineralized layers cooled progressively and subsolidus reactions occurred. The deposits were then to some extent recrystallized and remobilized during regional metamorphism, but from the late magmatic stage behaved essentially as closed systems with respect to PGE. From the proposed genetic model it follows that the presence of chromite layers in an basic-ultrabasic complex increases the prospecting potential for PGE, that silicate rocks above the upper chromite reef are an exploration target and that the absence of significant base-metal sulphides does not preclude the presence of Pt and Pd concentrations. The project was coordinated by the BRGM (France) and the partners were CNRS (France), the University of Copenhagen (Denmark) and the University of Manchester (United Kingdom).

 A back-propagation artificial neural network (ANN) model is proposed to discriminate zones of high mineral potential in the Rodalquilar gold field, south-east Spain. using remote sensing and mineral exploration data stored in a GIS database. A neural network model with three hidden units as selected by means of the k-fold cross-validation method. The trained network estimated a gold potential map efficiently. indicating that both preciously known and unknown potentially mineralized areas can be detected. These initial results suggest that ANN can be an effective tool for mineral exploration spatial data modelling.

 The platinum-group minerals (PGM) and chromite composition of the Zolotaya River gold placer, in the Russian Far East, indicate a genetic relationship to poorly exposed Late Permian dunite - hornblendite - gabbro intrusions of Uralian-Alaskan type. Ninety percent of the PGM grains studied consist of Pt-Fe alloy, typically with a rhodium content in the 1-2 wt.% range. Pt-bearing Os-Ir-Ru alloy occurs in three variants: (1) heterogeneous grains with an exsolution texture, (2) intergrowths of Os-Ir alloy with Pt-Fe alloy, and (3) homogeneous grains. Minerals of the erlichmanite-laurite series are relatively rare. Their composition varies from pure laurite through Os-bearing laurite to erlichmanite. Cooperite occurs in multiphase inclusions (in association with cuprorhodsite and chalcopyrite) in Pt-Fe alloy, and as a secondary rim (occasionally intergrown with platarsite and sperrylite) around Pt-Fe grains. Heterogeneous aggregates of Pt-Fe and Ru-Pt-Ir alloy have inclusions of arsenides of rhodium, iridium and ruthenium (cherepanovite, iridarsenite, ruthenarsenite) and a sulfarsenide-rich rim (irarsite).

 The composition and the associations of platinum-group minerals from the Salmon River placers, Goodnews Bay, Alaska, are compared with those of assemblages in other Alaskan-type complexes. Pt-Fe alloy, with an average value 25 at.% Fe, corresponds to isoferroplatinum (Pt3Fe). Ir concentrations in isoferroplatinum reach 15.4 wt.%, and Rh concentrations, as much as 2.27 wt.%. Tulameenite, which replaces the isoferroplatinum grains, appears as a rim with distinct boundaries. It forms a Pt2CuFe-PtFe series and also contains up to 2.64 wt.% Ni. Iridium and osmium occur as inclusions in the isoferroplatinum matrix and are depleted in Ru. The Ir content of osmium reaches 26 at.%, and the proportion of Os in iridium varies from 12 to 20 at.%. The concentration of Pt in iridium reaches 21.79 wt.%, which exceeds the limit of solubility of Pt in Ir at the maximum temperature of equilibration (similar to850degreesC). The composition of PGE-sulfarsenides corresponds to irarsite, with accessory platarsite and osarsite components: (Ir,Pt,Os)AsS. As predominates over S where substantial Pt exists, and excess S is present in irarsite with a high Os content. Solid solution along the erlichmanite (OsS2) - sperrylite (PtAs2) series is restricted. A wide range of compositions and levels of minor elements (Ir, Pd, Rh) in Pt-Fe alloys from different Alaska-type occurrences reflects geochemical features of the ore-forming system at the source. Fractionations 4 the PGE in the ore-forming system, from Ir-rich nuggets to Rh-rich and to Pd-rich alloys, reflect the decrease in temperature of the ore-forming system. Compositional variation and temperature range of alloy genesis decrease in the order from Goodnews Bay and Inagli to Galmoenan, Fifield, Nishnii Tagil, then to Tulameen and Yubdo, and down to Pustaya and Itchayvayam occurrences.

 Mineralization in the platinum-group elements occurs in lodes in the dunite - clinopyroxenite - gabbro Gal' moenan complex, in the Koryak-Kamchatka Platinum Belt of Russia, and in associated placers, namely those along the Levtyrinyvayam River and Ledyanoy and Penistiy creeks at different distances from the source. The grains of Pt-Fe alloy in lode deposits are isoferroplatinum and have Pt3Fe compositions, whereas those in the Levtyrinyvayam River placer range in composition front native to ferroan platinum and having iron contents generally higher than in the lodes (30-35 at.% Fe). Platinum grains from the Ledyanoy Creek placer are larger and more commonly intergrown with chromite than those from the Levtyrinyvayam River. Minor elements, which are indicators of conditions of formation of the Pt-Fe alloy, exhibit different concentrations in lodes and in the nearby placers. Platinum-iron alloy from the Levtyrinyvayam placer, the farthest from the Source, represents the eroded upper apical part of the intrusion. and contains mainly Pd. The Penistiy placer receives platinum from dunite and pyroxenite of a marginal facies of the complex. and thus platinum-iron alloy hearing Rh-Ir and Pd-Rh. Placer isoferroplatinum deposited in the Ledyanoy Creek, which drains dunite and chromite-bearing rocks in the south of the intrusion, is rich in It (LIP to 6.98 wt.%), as is the Pt-Fe alloy in the Gal'moenan complex (4.34 wt.% It). Most grains contain also a relatively high Rh content. Isoferroplatinum + iridium and isoferroplatinum + osmium are two main primary magmatic parageneses revealed in both placers and lodes, though the first is more commonly found in lodes and in the Ledyanoy Creek placer, and the latter in the Levtyrinyvayam River placer. Lodes mostly contain laurite, RuS2, and the Levtyrinyvayam placer bears all varieties of (Ru,OS)S-2, including Ru-free Rh-rich erlichmanite (Os,Rh)S-2. The irarsite - hollingworthite series predominates in lodes, and the irarsite - platarsite and osarsite - platarsite series predominate in the Levtyrinyvayam placer. Sperrylite in lodes may contain tip to 6.94 wt.% Sb. and thus differs from that in placers. The phase (Pt0.96Fe0.02)(Sigma0.98)(As1.17Sb0.75S0.09)(Sigma2), a member of the sperrylite - geversite solid solution, has been found in a lode. Complete solid-solution between RuS2 and RuAsS on one hand, and PtAsS and PtAs2, on the other, is demonstrated. In the Gal'moenan complex, concentrations of Ir decrease. and those of Pd and Os increase, toward the upper or apical parts of the complex.

 We have studied the platinum-group minerals (PGM) recovered from the Pustaya River placer deposit, situated within the Koryak-Kamchatka platinum-bearing belt of Alaskan-type intrusions in eastern Russia. The specific features of the geochemistry of the PGM of this placer deposit characterize the ore-forming system in the Alaskan-type source, which was partly eroded. The PGM concentrate (160 grains) consists of platinum nuggets, except for two grains, an iridium nugget and Pd-bearing gold (15.18% Pd). The platinum nuggets are represented by native platinum, isoferroplatinum and Pt-Fe alloy with an elevated concentration of Pd (up to 9.52 wt.%). Some nuggets are surrounded by rim's of PtCu, Pt3Cu, or PtS composition. Osmium, cooperite, vysotskite, platarsite, hollingworthite, irarsite, sperrylite, keithconnite, stibiopalladinite, rhodarsenide, vasilite, rhodian pyrrhotite and unnamed phases Pd2Te, (Pt,Pd)(3)S-2, (Pt,Pd,Fe,Cu)(3)S, and (Cu,Fe,Pd,Pt,Rh)(9)S-8 occur as inclusions in the platinum nuggets. The ore forming system is characterized by a depletion in Os, Ir, and Ru, and enrichment in Pd and Cu at the late-magmatic stage and Cu at the hydrothermal-metasomatic stage. Enrichment of the melt in Pd produced Pd-rich Pt-Fe alloys. Residual portions of the melt enriched in S, Pd and Cu filled gas vacuoles in platinum nuggets, in which the further crystallization of multi-phase parageneses [vysotskite, vasilite, (Pt,Pd)(3)S-2, (Pt,Pd,Fe,Cu)(3)S, (Cu,Fe,Pd,Pt,Rh,Ru)(9)S-8, bornite and chalcopyrite] occurred. Further hydrothermal-metasomatic transformations led to the appearance of hongshiite and cooperite rims on platinum.

 The zoned dunite-clinopyroxenite-gabbro massifs of the Platinum Bell of the Urals occur in an autochthonous manner in the accretion prism of the Ordovician-Silurian island-arc system. The massifs were formed above a subduction paleozone, yet outside of its main igneous rock belt (barrier zone). As follows from isotopic datings, the massifs might have formed in different periods of time in different segments of the Tagil paleoarc, during the junior and developed phase of its evolution. As follows from the geochemical data available, the most widely developed low-K subophitic gabbro norite and amphibole-pyroxene gabbro bear features of the most primitive island-arc tholeiite (Zr, Hf, U, Th, LREE), rift-zone tholeiites from active continental margins (P2O5, TiO2), and rare adakite magmas (Al2O3, Sr, Y). None of the volcanic rocks composing the Tagil paleoarc satisfy the requirements imposed upon volcanic rocks derived from the same magma at the same time in the same area. Similar paragenetic associations of dunite and clinopyroxenite, gabbro-norite and adakite granitoids, oceanic tholeiite and palingenetic rhyolite are known from the accretion prisms of the Pacific coasts of the Aleutian Islands, Alaska, California, Japan, and other regions. The mechanism of the formation of a chain of basic-ultrabasic massifs of the Platinum Belt is inferred to been associated with the subduction of a seismic ridge (active spreading zone), that is, with the functioning of "a window in the plate" under the accretion prism and the mantle-wedge base at the front of a paleoarc. The abnormally high heat-flow environment, the spreading impeded by the roof above the spreading zone, and the still-continuing asthenosphere upwelling were able to create conditions favorable for forced squeezing of the plastic mantle material into the newly formed forearc crust ("hot tectonics"), for the local melting of the hydrated oceanic plate and accretion-prism sediments, and for the fractionation and hybridism of heterogenic magma. The linear pattern of the chain of the massifs and their asynchronous intrusion can be explained by the intermittent sliding of a trench-seismic ridge-trench triple point along the convergent boundary. During the common subduction of a cold oceanic plate (Devonian active margin in the case of the Ural region), such anomalous magmatism does not take place in the fore-arc region.