This thesis reports the detailed study of 150 km2 of a newly discovered ophiolite terrane in the Chewore Inliers of northern Zimbabwe. It is apparently the oldest dated ophiolite (senso stricto) in Africa. Field and laboratory investigation of the Ophiolite Terrane (OT) shows that it comprises three distinct lithological groups. The Maunde Ophiolite Group is comprised of back - arc ophiolite - type assemblages, namely:-meta - mafic volcanics, sheeted meta - mafic dykes, meta - gabbro, meta - mafic cumulates and serpentinites. The Kaourera Island - Arc Group is comprised of a silica and potassium variable suite of extrusive volcanic lithologies which represent a tectonically excised island - arc complex. The Kadunguri Whiteschists are comprised of chemically restricted whiteschists which are characterised by the equilibrium assemblage of talc and kyanite and unusual yoderite + kyanite + chlorite + talc + dravite + hematite assemblages. The Ophiolite Terrane has had a single period of crustal thickening with a NNE - SSW shortening direction. This event is characterised by high shear strains with the development of non - coaxial intrafolial shear folds, an axial planar cleavage and an extension lineation which is parallel to the intrafolial fold hinges. Tectonic transport was directed obliquely toward the NNE. Associated clockwise metamorphism peaked within the upper amphibolite facies at 700°C and 10 kbar. The high pressure, moderate temperature Kadunguri Group whiteschists record peak metamorphic PT's of between 13 and 21 kbar and 550 - 650°C with the production of the second natural occurrence of the mineral yoderite. A plagiogranite sheet within the ophiolite has been SHRIMP dated at 1393 ± 22 Ma (Oliver et al., 1998) and is interpreted to be the age of the ophiolite. Investigation of the terranes to the north of the OT by Goscombe et al., (1998) reveal a previous, high T, low P, anti - clockwise, local granulite facies tectono - metamorphic cycle (M1) which is not recognised within the OT. All terranes are variably overprinted by the main clockwise PTt path, amphibolite facies, crustal thickening event (M2). The M1 cycle has been U / Pb SHRIMP dated from zircons within the Granulite Terrane at 943 ± 34 Ma (Goscombe et al., 1998) and interpreted to represent crustal extension associated with continental break up of Rodinia. Zircon overgrowths within the Granulite Terrane date the M2 cycle at 524 ± 16 Ma. The lack of M1 metamorphism and associated fabrics within the OT and the similarity in early M2 structures between all terranes indicate that juxtaposition of the Chewore terranes occurred during the earliest part of the M2 cycle i.e., at 524 ± 16 Ma. The plethora of Pan - African radiometric ages (c. 520 - 530 Ma) throughout the Zambezi Belt suggests that the belt containing the OT represents a complex east to west trending suture zone which divides West Gondwana into a 'northern' (Congo Craton) and 'southern' (Zimbabwe Craton) component. The Zimbabwe and Congo Cratons were therefore ultimately juxtaposed during the Pan - African Orogeny.

Apart from the concentrations of chromite in layers within the Great Dyke and other ultramafic complexes, chromite also occurs as interstitial grains throughout the olivine-bearing rock-types. These olivine-bearing rocks include no rites, gabbros, dunites and pyroxenites. Chromite concentration in these rocks varies from 0.48 to 3.09 per cent of the rock, usually in the form of chromite (Ahrens, 1965; Worst, 1960). A small fraction of this chromite settled to form chromitite layers whilst the remainder is retained within the rock mass as finely disseminated chromite and chromite interstitial to olivine. This retained chromite is much finer grained than layer chromite and is the primary source of eluvial chromite (Cotterill, 1981). During weathering of the serpentine rock and transportation by rainwater, the heavier chromite and magnetite grains are re-deposited along watercourses and vleis or valleys as the speed of the water is retarded sufficiently for the heavier particles to settle. The lighter serpentine material is removed and the chromite concentration in the soil is increased, thus resulting in eluvial chromite (Keech et ai, 1961; Worst, 1960; Prendergast, 1978). The concentration of chromite particles in soil can be up to 15 (or more) Cr₂O₃ %, resulting in economic and exploitable deposits, located primarily along the Great Dyke fiacks. A preliminary evaluation of the eluvials indicate that the Great Dyke could be host to up to 10 million tonnes of potential chromite concentrates which could be processed from such eluvial concentrates. These chromite-rich soils can be mined more cheaply than the traditional seams mining and processed into chromite concentrates through simple mechanical processing techniques of spirals, jigs and heavy media separators. The resultant chromite concentrates are of high quality and can be used to manufacture chromite ore briquettes, which are an alternative to lumpy chromite smelter feed. The main challenges to eluvial mining are the inevitable environmental degradation and coming up with methods that could possibly mitigate against such environmental damage. The distribution of these eluvials over vast plains as thin soil horizons, necessitate use of mobile concentrator plants and hence establishment of extensive infrastructure. These challenges, however, are not insurmountable and test mining and previous production runs have proved profitable. The eluvials are also associated with some lateritic nickel concentrations. The nickel occurs in close association with some oxide such as goethite and garnierite and is associated with iron-manganiferous soil pisolites. The analyses of these pisolites indicate high nickel grades of generally above 1.00 %Ni. Such high nickel-content of Great Dyke laterites warrant, further investigations.

A thesis submitted to the Faculty of Science at the University of the Witwatersrand, in fulfilment of the requirements for the degree of Doctor of Philosophy in Geology. Johannesburg 2017. / The Magondi Belt is a NE-trending Palaeoproterozoic mobile belt, composed of a succession of supracrustal metasediments and minor metavolcanics that is subdivided into the Deweras, Lomagundi and Piriwiri Groups. The Magondi Belt is located in north-west Zimbabwe and is bounded on its eastern flank by the Archaean Zimbabwe Craton and the Pan-African Zambezi Belt to the north. A connection between the Superior and Zimbabwe cratons has previously been made based on similarly aged dyke swarms across the two cratons. This matching magmatic barcode implies that the Superior and Zimbabwe cratons rifted away from one another circa 2.26 Ga based on the ages obtained for the Deweras lavas and the Chimbadzi Hill mafic intrusion. It was into this continental rift margin that the Magondi Supergroup sediments were deposited. The majority of the detrital and xenocrystic zircon ages from the Deweras Group are Archaean (2.86 to 2.63 Ga, with some inherited grains as old as 3.34 Ga); although a maximum depositional age of circa 2.29 Ga for the Deweras Group sedimentary rocks has been determined. Unconformably overlying these sediments, within an environment gradually transitioning from a passive margin into a back-arc basin environment, is the Lomagundi Group. These shallow marine sediments are then followed by those of the Piriwiri Group, deposited within a deeper water environment. Deposition of these two groups is constrained between 2.20 and 2.16 Ga, but may have continued up until the termination of the Magondi orogeny circa 1.99 Ga. According to the currently accepted model, the Magondi orogeny is the result of the Zimbabwe Craton colliding with an unknown continental mass, “Terra Incognita”, resulting in the formation of a Palaeoproterozoic Andean-type magmatic arc along the western margin of the Zimbabwe Craton (the arc is typified by the 2.06 - 2.02 Ga granites and gneisses of the Dete-Kamativi Inlier), which was subsequently thrust over the margin of the Zimbabwe Craton, the consequence of which was a Himalayan-style collision that resulted in high-grade metamorphism and the formation of collisional granitoids (e.g. the Hurungwe Granite) circa 1.99 Ga ago. It has also been established that the Dete-Kamativi Inlier, which flanks the western margin of the Zimbabwe Craton, is an extension of the Magondi Belt. Detrital zircons from paragneisses of the Malaputese Formation have ages ranging from 2.8 to 2.5 Ga, with the youngest grains constraining the maximum depositional age to be around 2.3 Ga. Thus, in terms of age and lithology, the correlation of the Malaputese Formation with the Deweras Group (maximum age of 2.29 Ga) is permissible. A south westward extension of 2.06 - 2.02 Ga granitoids – emplaced during the Magondi orogeny – is indicated by a number of localities in north-eastern Botswana and is believed to also be related to the Palaeoproterozoic magmatic arc. This study has recorded the first evidence of Archaean-aged basement within the Dete-Kamativi Inlier. Two orthogneisses with ages of 2.76 and 2.69 Ga provide strong evidence to suggest that the western margin of the Zimbabwe Craton may extend further to the west than previously recognised. It has also been confirmed, based on the recurrence of ~2.64 Ga aged zircons, in addition to older inherited grains ranging from 3.34 to 2.72 Ga, that the crust below the Magondi Belt is Archaean in age. This is not so, however, for the high-grade gneisses in the northern reaches of the Magondi Belt. It has been previously suggested that these supposed basement granites and gneisses represent an Archaean orogeny, but they are in fact Palaeoproterozoic in age, as represented by the syn-to-post-tectonic 2.02 Ga Hurungwe orthogneiss and the 1.95 Ga Kariba Granite. Additionally, a second, 1.96 Ga, orthogneiss contains zircons with younger metamorphic overgrowth rims that are Pan-African in age (545 Ma) and are attributed to the collision between the Kalahari and Congo cratons in the Neoproterozoic. It is therefore apparent that there is not enough evidence to support the existence of an Archaean “Hurungwe orogeny”. The Magondi orogeny was the heat source for a widespread mineralisation and metamorphic event between 2.15 and 2.03 Ga, based on titanite and apatite ages from samples of the Archaean Copper King and Copper Queen Domes within the Magondi Belt. There is also evidence of a second, younger, mineralisation event, which primarily affected both the Dete-Kamativi Inlier and the Choma-Kalomo Block (south east Zambia). U-Pb data on columbite-tantalite grains (corroborated by 40Ar-39Ar dating of mica separates) from tin-bearing pegmatites within both the Choma-Kalomo Block and the Dete-Kamativi Inlier indicates that mineralisation occurred simultaneously within these two terranes between 1.06 and 0.98 Ga. The similarities (particularly with regards to mineralisation), between the Choma-Kalomo Block and the Dete-Kamativi Inlier imply that these two terranes had a shared history, potentially as far back as the Palaeoproterozoic, but were certainly juxtaposed by 1.06 Ga when the pegmatites were emplaced. The previously undated metasediments of the Choma-Kalomo Block have revealed an abundant Palaeoproterozoic component (2.04 - 1.86 Ga), contradicting the prevailing understanding that the Choma-Kalomo Block is solely Mesoproterozoic in age (on account of the granitoids, which were previously dated at 1.37 and 1.18 Ga). The Choma-Kalomo Block was also thought to constitute an exotic terrane with respect to the neighbouring Dete-Kamativi Inlier and Archaean Zimbabwe Craton. Based on the geochronology presented here, a new model is proposed whereby the thinner lithosphere beneath the Choma-Kalomo Block is either a primary feature or one that resulted from subduction erosion and delamination processes associated with the formation of multiple continental margin magmatic arcs. / MT 2017

From Introduction: The object of the investigation was to examine the variations in the granite and the relation between the granite and the surrounding metamorphosed sedimentary rocks of the Shamvaian System (Primitive). Although this object may not have been fully achieved, it has been possible to show that a wide variety of rocks is present in the area, and a few suggestions as to their origin have been made.

This thesis focuses on the petrology and geochemistry of the lower Pyroxenite Succession of the Great Dyke of Zimbabwe in an area to the south ofMutorashanga. Particular emphasis is placed upon the economically important chromitite C5, and on the pervasive serpentinization of olivinerich rocks. An overview of the Great Dyke, including the Satellite Dykes, the structure and stratigraphy of the Great Dyke, the economic resources of the Great Dyke, and the evolution of the Great Dyke magma, is given. A review of the geodynamic history of the Zimbabwe Archaean craton, which culminated in widespread cratonisation and emplacement of the Great Dyke is also provided. The silicate rocks of the lower Pyroxenite Succession are highly adcumulate dunites and orthopyroxenites, with well-developed granular textures and a restricted mineral assemblage of olivine and pyroxene, with very minor plagioclase and clinopyroxene. Within cyclic units, the silicate rocks commonly display a textural and modal progression from granular dunite through poikilitic harzburgite, granular harzburgite, and olivine orthopyroxenite, to granular orthopyroxenite. Chromitites commonly occur at the base of each cyclic unit, these are thin, massive, coarse-grained layers, and are shown to be modified, texturally and compositionally, by postcumulus annealing processes. The olivine-rich rocks are pervasively serpentinized to a depth of over 300 metres. The serpentites typically display well-developed pseudomorphic mesh textures, with a slight overprint of nonpseudomorphic interpenetrating textures and late-stage cross-cutting veins. X-Ray diffraction studies indicate that chrysotile is the dominant serpentine mineral, and also reveal the presence of a nickeliferous magnesium hydroxide, occurring as an intimate admixture with serpentine, and believed to be a nickel-bearing analogue of brucite. Mineral and whole rock compositions of chromitite and silicate rocks highlight the strongly magnesian nature of the Ultramafic Sequence. Studies ofthe footwall chromites below chromitite C5 are consistant with a model of replenishment of primitive magma into the Great Dyke magma chamber, at the base of each cyclic unit. The magma injection and subsequent mixing with the evolved resident magma gives rise to chromitite fonnation, and a causes a reversal of the fractionation trend, resulting in a return to more primitive compositions in the silicate rocks. The silicates display an overall fractionation trend that reflects the evolving composition of the parental magma.

Gold exploration was conducted in northern margin, granulite-facies rocks of the Limpopo Belt. Methods used in the prospecting include drainage, soil and rock geochemistry, geophysical surveys, geological mapping, trenching and diamond drilling. These techniques successfully led to the discovery of two medium size, mesothermal gold deposits (Grid 2s and Grid 4). Objectives of this study were to (a) document the exploration methodology used; (b) describe the regional geology; (c) establish a mineral deposit model; (d) outline the methods and results of various exploration techniques; (e) outline follow-up procedures and evaluation of anomalies; and (f) discuss results of the exploration exercise and conclusions. The granulite-facies terrain comprises Charno-enderbites, mafic and felsic to intermediate metavolcanic rocks and meta-sediments. Renco Mine situated immediately east of the study area, was selected as the ore deposit model for the exploration program. Gold mineralization occurs in shear and thrust zones within an enderbite. The gold deposits are structurally controlled by a first-order, Sinistral transcrustal Mauch Shear Zone, which is parallel to a regional east-northeast penetrative foliation. The deposits are in dilation zones where the Mauch Shear (a) is intersected by a dextral east-west shear (Grid 2s), or (b) has a sinistral splay (Grid 4 and Renco). Close to these deposits, the Mauch Shear is in contact with a "greenstone belt", which is a possible source of crustal metamorphic ore fluids and gold. The Grid 2s deposit contains fine-grained, disseminated free gold, and small amounts of pyrrhotite, pyrite and chalcopyrite in quartz veins within third-order shears in K-feldspar granite. K-feldspar, sericitic, silicic, sulphidation and carbonate alteration characterizes the deposit, which has a proposed mantle-degassing model. The Grid 4 deposit is magmatic porphyry-type, with CuMo and Au in third- and fourth-order shears respectively. Mineralization comprises disseminated to semi-massive pyrrhotite, pyrite, chalcopyrite, sphalerite, bismuth, molybdenite and gold. Wall rock alteration includes biotitic, chloritic, silicic, sulphidation and carbonate. In Grid 2s, Grid 4 and Renco deposits, the alteration mineral assemblages are in three facies, which are granulite, amphibolte and greenschist. In the three deposits, the mineralization occurs with the amphibolite-facies, indicating post-peak, retrograde metamorphic conditions. / Illustrations (maps) only available in print form at Cory Library / KMBT_363 / Adobe Acrobat 9.54 Paper Capture Plug-in

A remarkable correlation between hydrothermal gold mineralisation and geological structure is discussed. The mineralisation occurs in vein quartz which occupies a fissure in Archaean greenstones of the Basement Complex in Rhodesia, It has been determined that gold which is silver-rich is typical of low-grade ore, and is associated in space with sulphides that crystallised early in paragenesis. Silver-poor gold is characteristic of highgrade ore which is concentrated along the crestal zone of anticlinal warps in the fissure. It is suggested that the local pattern of fracturing and folding is related to fundamental wrench faulting.

The geology of the Shamrocke area is described relative to its regional setting and position within the stratigraphic succession of the Lomagundi System. The stratigraphy in the immediate vicinity of the Shamrocke Mine is detailed and discussed relative to the work of others south of the project area and in other regions. The petrography of the rocks of the Shamrocke Mine area is described and the results of a great deal of mineralogical work recorded. Maps of the project area are presented at various scales from field and photogeological evidence. The thesis area is situated on the South Zambezi Escarpment of Rhodesia, and the geology described included the basaI succession of the Lomagundi System and the pre-Lomagundi Escarpment Series. The Shamrocke Mine is located on a copper orebody associated with a granulite or granofels zone within the Dolomite Series of the Lomagundi System. This ore zone granulite appears to be a metasomatised calcareous grit some 1000 feet above the upper contact of the Deweras Series (basal Lomagundi) and, within the graphitic schist and phyllite, below a dolomitic horizon in the Dolomite Series. The Lomagundi succession in the Mine area unconformably overlies the pre-Lomagundi gneiss and meta-arkose of the older, metamorphosed and deformed Escarpment Series. The basal meta-arkose, meta-quartzite and coarse schist of the Deweras Series ascends southwards through the Dolomite Series (graphitic phyllite and schist, granulite, calcareous grit, dolomite, limestone), and the Argillaceous Series (schist, phyllite, quartzite), the beds dipping steeply to the south at an angle of between 50º and 70º. Post-Lomagundi plagioclase amphibolite (altered, intrusive meta-diabase) forms Iarge semi-concordant and transgressive sills throughout the area, particularly along the contact between the Deweras and Dolomite Series. The Shamrocke Mine is on the northern limb of a large synclinal structure, the Rusere Syncline, which forms a large embayment of Lomagundi rocks into the pre-Lomagundi gneisses and granodiorites northeast from the Mine. The fold is overturned to the east and southeast. The copper mineralisation within the area and in the areas to the south is considered to be invariably associated with the basal rocks of the Lomagundi System. It occurs within both the Deweras and Dolomite Series rocks and more often than not Iies close to the contact between these two Series. The sulphide mineralisation of the Shamrocke orebody is considered, from the results of the present study, to be metasomatically emplaced during carbonate metasomatism, either from an extraneous source or from within the ore zone rock itself. The present writer favours the origin of the copper sulphide to be original syngenetic sulphide of the basal rocks of the Lomagundi depository, which has been mobilised and metasomatically relocated, possibly by the effects of regional metamorphism related to intense deformation. It is perhaps not fortuitous that the majority of the copper occurrences in the area occur where the basaI beds of the succession have been cross-folded. The copper ore comprises a simple suite of minerals, the main constituents being chalcopyrite, cubanite and pyrrhotite. The deposit is compared relative to the other copper deposits of the Lomagundi System.

[Preface] The detailed geological survey of the country described in this bulletin commenced in 1953 and was completed in 1956 with the mapping of 1,480 square miles of country. Information is supplied on 227 mines which produced mica alone or both mica and beryl, and 134 mines which produced only beryl. In addition, wolfram, gold, graphite, kyanite and rutile deposits and mines are described. Mineral outputs are given up to end of 1959. The field mapping and writing of the bulletin is entirely the work of Mr. Wiles. Mapping of the Field has established that, with a few exceptions, a metamorphic control has operated in the development of economic sheet mica and has led to the conclusion that very many of these pegmatites are composite bodies which have derived their material from two sources: one metamorphic and the other igneous. This conclusion was arrived at after a detailed petrographic investigation of the metamorphic rocks. The map at the end of this bulletin was drawn by Mr.A. H. Barrie and the diagrams are largely the work of Mr. D. O. L. Levy. Chemical analyses are by Messrs. A. J. Radford and E. Golding and were made in the Geological Survey Laboratory. Some of the photographs are by H. J. Cotterel of the Geological Survey while the remainder were taken by the Planning and Technical Services of the Federal Ministry of Home Affairs (formerly the Federal Information Department). It gives me pleasure to acknowledge the assistance and facilities offered the Department by mining men and residents in the district.

The Archean (~3.0 Ga) Buhwa Greenstone Belt, and surrounding granite-gneiss terrane, is the least understood major greenstone belt in the Archean Zimbabwe Craton, despite occupying a critical position between an early Archean continental nucleus and the Limpopo Belt. The cover succession in the Buhwa Greenstone Belt, which was probably deposited on the margin of this nucleus, is divisible into shelfal and basinal facies associations separated by a transitional facies association. The shelfal association consists mostly of quartzarenite and shale, but also contains a thick succession of iron-formation. Geochemical characteristics of the shales indicate that the source terrane consisted of several lithologies including tonalite, mafic-ultramafic volcanic rocks, and granite that underwent intense chemical weathering. Basinal deposits consist dominantly of greenstones, with less abundant chert and ironformation. The cover succession, which was deposited on a stable shelf transitional to deep water, has no stratigraphic equivalents elsewhere on the Archean Zimbabwe Craton. However, time and lithologic correlatives in the central zone of the Limpopo The Archean (-3.0 Ga) Buhwa Greenstone Belt, and surrounding granite-gneiss terrane, is the least understood major greenstone belt in the Archean Zimbabwe Craton, despite occupying a critical position between an early Archean continental nucleus and the Limpopo Belt. The cover succession in the Buhwa Greenstone Belt, which was probably deposited on the margin of this nucleus, is divisible into shelfal and basinal facies associations separated by a transitional facies association. The shelfal association consists mostly of quartzarenite and shale, but also contains a thick succession of iron-formation. Geochemical characteristics of the shales indicate that the source terrane consisted of several lithologies including tonalite, mafic-ultramafic volcanic rocks, and granite that underwent intense chemical weathering. Basinal deposits consist dominantly of greenstones, with less abundant chert and ironformation. The cover succession, which was deposited on a stable shelf transitional to deep water, has no stratigraphic equivalents elsewhere on the Archean Zimbabwe Craton. However, time and lithologic correlatives in the central zone of the Limpopo ~2.9 Ga in southern Africa. At ~2.9 Ga, the northern margin of the greenstone belt experienced kilometerscale, oblique-slip dextral shearing. This shear zone and the surrounding margins of the greenstone belt were later intruded by the ~2.9 Ga Chipinda batholith, which ranges from granitic to tonalitic in composition. A number of events occurred during the time period spanning 2.9-2.5 Ga and current geochronology cannot separate their order; some are known to be coeval. Crustal shortening to the northwest, which resulted in map-scale folding of the cover succession (and surrounding batholith) and greenschist-facies metamorphism, occurred along a set of discrete high-angle reverse-sense shear zones in response to uplift the Northern Marginal Zone of the Limpopo Belt over the Zimbabwe Craton. Two suites of potassic granites were intruded into the area near the end of reverse shearing. Analysis of a conjugate fault pair that is developed within one of the potassic granite suites, yields a principal compressive stress consistent with continued northwest-directed crustal shortening. The region was stabilized by ~2.5 Ga, with intrusion of the Great Dyke of Zimbabwe. It is possible that the last events to affect the area, which include sinistral shearing, transecting cleavage development, and northwest-striking open folding, took place during the 2.9-2.5 Ga time intervaL These structures post-date regional folding and metamorphism, but because of limited magnitude and extent, do not show obvious cross-cutting relationships with other rocks or structures. A tenable alternative is that these late structures formed at ~2.0 Ga. an age that is proving to be of great significance in the evolution of the Limpopo Belt and along parts of the southern margin of the Zimbabwe Craton. / Ph. D.