Abstract The gold-rich Superior, Canada, and Yilgarn, Australia, cratons have similar geologic histories dating back to the Mesoarchean and showing strong parallels in the Neoarchean. Orogenesis in each craton is marked by a shift from dominant volcanism to dominant clastic sedimentation above unconformities, followed by granitic plutonism, progressive deformation, and dynamothermal metamorphism. The terminal stages of orogenesis correspond to the intervals of 2660 to 2650 Ma in the Superior craton and 2660 to 2630 Ma in the Yilgarn cra ton. The Yilgarn and Superior cratons contain an estimated 9,200 and 8,500 t Au, respectively. Most of the sig nificant gold deposits (>100 t Au) are concentrated in a few narrow, highly endowed gold belts along which the deposits cluster into camps, commonly spaced every 30 to 50 km. Gold deposits of both cratons show similar tonnages and grades, and their size distributions define a Pareto, rather than log-normal distribution. Large deposits are rare but account for most of the gold endowment of each craton. Three recurring host-rock associations account for a majority of large deposits: iron-rich mafic igneous rocks, iron-rich sedimentary rocks, and felsic to intermediate porphyry stocks and dikes. Most deposits, particularly large ones, occur in greenschist-grade rocks and are associated with shear zones, faults, or folds. Gold miner alization styles include quartz-carbonate veins, sulfidic replacements in banded iron formation (BIF), crusti form carbonate-quartz veins and associated sulfidic replacement lodes, disseminated-stockwork zones, sulfide rich veins and veinlet zones, and massive sulfide lenses. Wall-rock alteration assemblages vary with mineralization style and metamorphic grade. Most deposits consist of a single style of mineralization, but many of the large ones combine two or more of these, and some large deposits are unique in their metal associations. The diversity of styles of mineralization, wall-rock alteration assemblages, and overprinting relationships re quire more than one episode of gold mineralization and more than one ore-forming process. Geologic parage neses, coupled with isotopic age constraints, show that, although the Archean histories of both cratons span >300]m.y., the majority of gold deposits formed during the final 30 to 50 m.y. of that time span, corresponding to the orogenic phase. The majority of gold deposits can thus be regarded as orogenic in timing, but with the available constraints clearly pointing to the existence of more than one mineralizing event and involving dif ferent mineralization types and processes. The best-endowed gold camps (Timmins and Red Lake in the Superior craton; Kalgoorlie, Granny-Wallaby, and Sunrise Dam in the Yilgarn craton) commonly possess an anticlinorial structure, komatiitic and basaltic rocks in the core giving way to stratigraphically higher volcanic and clastic sedimentary rock units. Such camps are further marked by coarse clastic rocks deposited above the metavolcanic rock sequences, by concentrations of shallow-level porphyritic intrusions, by extensive carbonate alteration, by multiple styles and ages of gold mineralization and, in most, by through-going regional faults. However, these characteristics are also shared by a number of less-endowed gold camps. The best-endowed gold belts are distinguished by substantial volumes of komatiite, by a high degree of preservation of supracrustal rocks, by structural highs that juxtapose the lower and uppermost parts of the stratigraphic column, by multiple styles and ages of gold mineralization, and by world-class deposits of other metals. The gold belts commonly are aligned along crustal-scale faults that rep resent long-lived structures, which acted as crustal-scale magma and fluid conduits and also influenced coarse clastic sedimentation. Abundant komatiites may reflect the first tangible connection to the deep crust and mantle, but the nature of the subvolcanic crust, ensimatic in the Timmins-Val d’Or, Superior craton, and ensialic in the Wiluna-Norseman belts, Yilgarn cration, seems unimportant in determining gold prospectivity. Significant uncertainty remains concerning the timing of formation of the deposits, the models that best explain their characteristics, and the fundamental causes of the high concentration of gold in a few areas. Various models have been proposed, invoking volcanic, magmatic, and orogenic (metamorphism and/or deformation) processes. The synorogenic model best accounts for the Au-only quartz-carbonate veins and temporally related mineralization styles. However, synvolcanic and magmatic hydrothermal models are also required to explain the presence of Au base metal deposits and those deposits overprinted by significant deformation and meta morphism. The specific histories of the gold belts and the known constraints on timing of deposits suggest that all of these processes have contributed to the gold endowment, although it is difficult to separate orogenic from magmatic processes because they closely overlap in time and space. Despite the presence of synorogenic quartz-carbonate veins throughout the greenstone belts of both cratons, large deposits of this style are mainly restricted to the gold belts, where they also coexist with large deposits of other styles of gold mineralization. The presence of multiple ages and styles of gold mineralization in the best-endowed gold belts indicate a unique locus of successive formation of gold deposits from various processes operating at different stages of the orogenic phase of the evolution of these belts. This would explain the common overprinting of the early deposit types, potentially of synvolcanic or synplutonic origin, by syn orogenic ones. The concentration of multiple types and ages of significant gold deposits in well-defined gold belts is not a unique feature of the Superior and Yilgarn cratons but is shared by Tertiary gold belts of Nevada, such as the Walker Lane, the Battle Mountain-Eureka trend, and the Carlin trend. This must be a reflection of fundamental crustal structure, and perhaps composition of subcrustal mantle, as much as local ore-forming hydrothermal processes.

Abstract The Abitibi greenstone belt, the largest and best-preserved Archean granite-greenstone complex in the world, consists of five major lithologic assemblages which formed in four distinctive geotectonic environments. These comprise: (1) a tholeiitic and komatiitic assemblage formed in an oceanic extensional environment; (2) a calc-alkalic volcanic and volcanic-derived sedimentary rock assemblage formed in an island-arc environment; (3) an assemblage of craton-derived quartzose sedimentary rocks with interbedded komatiitic volcanic rocks formed on a passive continental margin; and (4) an assemblage of molasse-type sedimentary rocks, alkalic volcanic rocks, and felsic intrusions. The oceanic, arc, and continental margin assemblages were imbricated and tectonically stacked during the collision of a northward-moving continent and its attached marginal sediments with the arc, as a result of subduction of the intervening ocean crust. The molasse assemblage formed along the collisional suture zone. Interbedded alkalic volcanic rocks, syn- to postdeformational felsic intrusions, and the CO 2 -rich hydrothermal fluids responsible for widespread hydrothermal alteration and gold mineralization, all broadly part of this molasse assemblage, were generated by linked lower crust-mantle processes. The effect of the high density of the oceanic rocks on pressure-temperature conditions at the base of the tectonically thickened crust may have been instrumental in triggering the processes which resulted in auriferous fluids. Anatectic melts and gold ore-forming hydrothermal fluids were generated at depth as an end-stage manifestation of the collisional event. Rocks of the oceanic assemblage were the possible source of the gold. This model provides a rational and internally consistent explanation for the association of gold mineralization with greenstone belts, major faults, molasse-type sedimentary rock assemblages, felsic porphyry intrusions, and widespread hydrothermal alteration. Also, the model is consistent with the late timing of the mineralization in the geologic development of the Abitibi belt in particular, and both Precambrian and Phanerozoic greenstone belts in general.