• journal_title：The Canadian Mineralogist
• Contributor：R. Grant Cawthorn ; Christopher A. Lee ; Robert P. Schouwstra ; Peter Mellowship
• Publisher：Mineralogical Association of Canada
• Date：2002-
• Format：text/html
• Language：en
• Identifier：10.2113/gscanmin.40.2.311
• journal_abbrev：Can Mineral
• issn：0008-4476
• volume：40
• issue：2
• firstpage：311
• section：Platinum-Group Minerals: Ore Mineralogy

The abundances of the platinum-group elements (PGE), and their relative proportions, are summarized for all the mines exploiting the Merensky Reef on both western and eastern limbs of the Bushveld Complex, in South Africa. Grades for the mined reef interval are uniform over 100 km of strike length, at 5–7 g/t, except for Union mine, where it is currently 8 g/t. The relative proportions of the various PGE are extremely constant over this same distance. In contrast, the platinum-group minerals (PGM) vary enormously. On a regional scale around the western limb, the mineralogy ranges from alloy-dominated in the northwest to sulfide- or telluride–arsenide-dominant in the southeast. Local features that are important in terms of mining are potholes and discordant ultramafic bodies. Comparative studies of the PGE from close to and within such structures again show constant absolute and relative abundances. Limited data on the PGE from PGE-mineralized pipes (Onverwacht, Mooihoek and Driekop) suggest that they are Pt-dominated, and have a totally different relative pattern of distribution of the PGE than that in the stratiform mineralization. The PGM in the vicinity of these pipes also tends to be dominated by alloys, and in potholes by alloys and tellurides and arsenides. Hence, on both a regional scale and very local scale associated with disturbances in the reef, there is a dichotomy between the constancy of the PGE and the variability of the PGM. These observations suggest that the primary mechanism concentrating the PGE is a first-order process, and that the evolution of the PGM is the result of secondary processes related to cooling, local changes in f(S₂) in the crystallization environment, and subsolidus re-equilibration. The debate as to whether the PGE are derived by fluid scavenging of a footwall succession or have accumulated from supernatant magma is reexamined, and unresolved problems pertaining to both processes are highlighted.