Thesis (PhD)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: The discovery of lucrative diamond deposits along the west coast of Southern Africa about 1200 kilometres from the Kimberley region during the period 1908 to 1927, gave rise to a number of different theories with respect to their possible provenance. These included the transportation of diamonds from unknown sources in southern Namibia by south-flowing rivers, hidden on- and off-shore kimberlites along the coast, and transportation by west-bound rivers from the hinterland. Subsequent research has shown that the latter is the only plausible theory. The discovery of marine and coastal diamond deposits as far south as the Olifants River estuary showed that the Vaal-Orange drainage in its current form could not have been the only conduit for diamonds to the coast, and the drainage evolution of southern Africa was interpreted as comprising essentially the following two main palaeo-fluvial systems active in the formation of the world's only known diamond mega-placer deposit:  The Karoo River with its headwaters similar to those of the modern Orange and Vaal Rivers and entering the Atlantic Ocean via the present-day Olifants River;  The Kalahari River that drained southern Botswana and followed the route of the modern-day Molopo River, entering the Atlantic Ocean in the vicinity of the present Orange River mouth. An important shortcoming of the above model is that it could not account for the fact that diamond distribution along the west coast shows a marked increase in grade and average stone size at the estuaries of all the major rivers draining from the escarpment to the Atlantic between the Olifants and the Orange Rivers. The presence of fluvial diamond deposits along the courses of the Buffels, Swartlintjies, Spoeg, Horees and Groen Rivers confirms that the increased grade and diamond size at their estuaries is not a function of large bays and rougher bottom topography associated with the rivers, although these could have contributed to this phenomenon. This proves that the catchments of the rivers between the Olifants and Orange Rivers also had access to diamondiferous debris, although they were not in contact with these two major drainages. A number of researchers proposed that diamonds liberated from pre-Karoo kimberlites were moved from their primary hosts to the south-western parts of the subcontinent by Dwyka glacials. From the above it is clear that nearly a century after the discovery of diamonds along the west coast of southern Africa consensus regarding their origin had not been reached. The aim of this study was therefore to establish a model explaining the most likely sources and distribution history of the more important alluvial diamond deposits in southern Africa. The methodology comprised a study of 1878 diamonds collected from 25 alluvial and two kimberlitic sources for comparison with known similar data from 12 kimberlitic populations in southern Africa. The diamond study was supplemented by a study of sedimentary clasts from bulk gravel samples taken along the Middle and Lower Orange River as well as Scanning Electron-microscope (SEM) Analyses of garnet grains and zircon geochronology. The evidence from the study does not support the postulated existence of a former Karoo River. The surface features of diamonds, notably brown spots indicating – in the context of southern Africa - liberation from pre-Karoo kimberlites, as well as the results of Fourier Transform Infrared analyses revealed that the populations at Kwaggaskop along the Sout River, previously considered an erosion remnant of the Lower Karoo River and those occurring south of Brandvlei and Van Wyksvlei in the valley of the Sak River, previously considered to have been reworked from the Middle Karoo River, differ profoundly from each other. In addition, the surface feature studies and Fourier Transform Infrared Analyses clearly show major distinctions between the diamond populations from the Sout River-Olifants River estuary and those from the Kimberley kimberlite province which was said to have supplied diamonds in large quantities to the Olifants River estuary via the postulated Karoo River. Furthermore the idea of a palaeo-Gamoep River playing a significant role in the transportation of diamonds to the west coast is favoured by the presence of brown-spotted diamonds and diamonds with Platelet Preservation Indices revealing severe platelet destruction that could be traced through Bosluispan in the Koa River valley, the Buffels River valley, the Buffels River estuary and to the shallow marine environment north of the Buffels River. Zircon geochronology confirmed the role of the Orange River in the denudation of the sub-continent. With respect to the drainage evolution and diamond distribution in southern Africa the results of this study indicate a complex diamond dispersal model that differs in some respects from prevailing theories. It shows that diamonds liberated from pre- Karoo kimberlites in the north-eastern part of the sub-continent were initially moved in a south-westerly direction by pre-Karoo drainages, then by Dwyka glaciers and ice sheets. Ultimately, after liberation from exhumed glacial and fluvial deposits and together with diamonds subsequently liberated from Jurassic and Cretaceous kimberlites, Cretaceous and younger drainages provided the transport toward the Atlantic Ocean where the diamonds were concentrated along shorelines and in bedrock trap sites. Significant quantities did not reach the coast, but were locked up in fluvial sediments in erosion remnants like terraces, karstic depressions and other segments of palaeo-channels along the way. The presence of diamonds with FTIR characteristics reminiscent of those from Orapa and Jwaneng in the Orange River deposits as well as in a raised marine terrace in southern Namaqualand and in marine deposits north of Concession 12A, also negates the possible existence of a palaeo-Kalahari River, unless it was a very young system that did not interrupt the south-bound dispersal of Botswana diamonds during the Late Oligocene-Early Miocene. The study also included microscopic examination of a parcel of diamonds from the enigmatic Skeleton Coast deposits, north-western Namibia. These results confirmed the conclusion based on geological and geomorphic grounds that these diamonds cannot be linked to the Oranjemund deposits, while their surface features showed that pre-Karoo sources comprise the most likely provenance for the Skeleton Coast diamonds. Thus the combination of FTIR analyses and surface feature studies of diamonds, zircon geochronology and SEM analyses of garnets allowed the formulation of a revised model for the distribution of alluvial diamonds and the drainage history of the sub-continent since the Middle Cretaceous, while the study of sedimentary clasts confirmed the repeated occurrence of high energy fluvial conditions – especially evident in the palaeo-Orange River sediments – that contributed to the high percentage of gem stones in the surviving alluvial diamond populations due to the destruction of poor quality diamonds. / AFRIKAANSE OPSOMMING: Die ontdekking van ryk alluviale diamantafsettings aan die suider-Afrikaanse weskus, meer as 1200 kilometer van die Kimberley-omgewing af tussen 1908 en 1927, het 'n aantal teorieë omtrent moontlike provenansgebiede vir hierdie afsettings tot gevolg gehad. Dit het gewissel van die suidwaartse vervoer van diamante vanaf bronne in suidelike Namibië, diamantdraende kimberliete in die kusvlaktes of op die vastelandstoep onder huidige seevlak, tot die vervoer van diamante deur weswaarts-vloeiende riviere vanuit die binneland. Geen ontdekkings wat eersgenoemde teorie kon ondersteun is in Namibië gemaak nie. Verder, namate meer gevorderde navorsingsresultate aan die lig gekom het, het dit duidelik geword dat kimberliete wat weg van 'n antieke kraton geleë is, grootliks sonder diamante is, en gevolglik het die idee van nabygeleë diamantdraende kimberliete in die kusvlakte of op die seebodem as bron, onaanvaarbaar geword. Grootskaalse wes- tot suidweswaartse vervoer van diamante het gevolglik die enigste aanvaarbare alternatief gebied. Die ontdekkiing van aan- en aflandige mariene afsettings tot so ver as suid van die Olifantrsrivier het getoon dat die Vaal-Oranjestelsel in sy huidige vorm nie die enigste vervoerkanaal vir diamante na die weskus kon wees nie. Die dreineringsgeskiedenis van suidelike Afrika was gevolglik vertolk aan die hand van twee voorgestelde groot oer-rivierstelsels, naamlik: - Die Karoorivier met sy bolope naastenby soortgelyk aan dié van die moderne Oranje- en Vaalriviere, en wat langs die huidige Olifantsrivier uitgemond het; - Die Kalaharirivier wat die suide van Botswana gedreineer het, en min of meer die roete van die huidige Moloporivier gevolg het, met sy monding baie naby aan dié van die moderne Oranjerivier. 'n Belangrike tekortkoming in bogenoemde model is die feit dat dit nie 'n verduideliking bied vir die volgende feit nie: Diamant-produksiedata van die Suid-Afrikaanse weskus toon 'n skielike toename in graad (karaat per 100 ton) en gemiddelde steengrootte van diamante by die monding van al die belangrike riviere tussen die Olifants- en Oranjeriviere, wat vanaf die platorand na die Atlantiese Oseaan dreineer. Die feit dat fluviale diamantvoorkomste in die valleie van die Bufffels-, Swartlintjies-, Spoeg-, Horees- en Groenriviere aangetref word, bevestig dat hierdie verskynsel nie net aan die teenwoordigheid van kus-inhamme en ruwer vloertopografie wat met die riviermondings geassosiëer is, toegeskryf kan word nie, alhoewel dit wel „n bydrae tot hierdie waarneming kon maak. Dit bevestig dat hierdie riviere wel in hul opvang-gebiede ook toegang tot diamanthoudende puin gehad het, sonder enige kontak met die Olifants- of Oranjeriviere. 'n Aantal navorsers het die gedagte geopper dat diamante wat uit voor-Karoo kimberliete vrygestel was, deur bewegende ysplate en/of gletsers vanaf hul provenansgebiede na die suidweste van die subkontinent vervoer is. Uit die voorafgaande paragrawe is dit duidelik dat, ongeveer ʼn eeu ná die ontdekking van diamante langs die suider-Afrikaanse weskus, daar nog nie eenstemmigheid bereik is oor die oorsprong van hierdie diamante nie. Die doel van hierdie studie was gevolglik die daarstelling van „n model wat „n aanvaarbare verduideliking bied vir die verspreiding en afsetting van sommige voorkomste van spoeldiamante in suidelike Afrika soos tans waargeneem. Vir hierdie doel is 1878 diamante afkomstig vanuit 25 alluviale en twee kimberlietvoorkomste ondersoek. Die resultate is vergelyk met soortgelyke inligting wat bekend is vir diamantpopulasies vanuit 12 suider-Afrikaanse kimberliete. Die diamantstudie is aangevul met die ondersoek van spoelklippe vanuit gruismonsters wat langs die Middel- en Benede Oranjerivier versamel is asook Skanderings-elektron Mikroskoop-analises (SEM) van granaatkorrels en sirkoon-geokronologie. Die resultate van hierdie studie ondersteun nie die hipotese van „n eertydse Karoorivier nie. Die teenwoordigheid van bruin spikkels op diamante wat – in die konteks van die geologiese geskiedenis van suidelike Afrika – vrystelling vanuit vóór- Karoo kimberliete impliseer, asook die resultate van FTIR-analises dui op „n komplekse model wat „n alternatief bied vir bestaande sienswyses. Dit toon dat die diamantpopulasies by Kwaggaskop langs die Soutrivier wat veronderstel was om die Benede Karoorivier te verteenwoordig, en dié wat suid van Brandvlei en Van Wyksvlei in die vallei van die Sakrivier aangetref word en veronderstel was om afkomstig te wees uit die Middel Karoorivier, drasties van mekaar verskil. Dit openbaar ook beduidende verskille tussen die diamantpopulasies van die Olifantsriviermonding en dié van die Kimberley-omgewing waarvandaan die veronderstelde Karoorivier groot hoeveelhede diamante aan die Sout-Olifantsrivier sou gelewer het. Verder verskaf die teenwoordigheid van diamante met bruin spikkels en diamante met eienskappe wat toon dat hul stikstofplaatjies vernietig is, „n skakel tussen Bosluispan in die vallei van die Koarivier en die seegebied noord van die Buffelsrivier, via die Buffelsriviervallei en die Buffelsriviermonding, en hierdie feite ondersteun gevolglik eerder die voorstel dat groot hoeveelhede diamante deur die paleo-Gamoeprivier na die weskus vervoer is. Die teenwoordigheid van diamante met FTIR-kenmerke soortgelyk aan dié van Orapa en Jwaneng in die Mid-Oranje afsettings, 'n mariene terras in die suide van Namakwaland en in mariene konsessies noord van Seegebied 12A, opponeer ook die gedagte van 'n paleo-Kalaharirivier, tensy laasgenoemde 'n baie jong stelsel was wat nie die suidwaartse beweging van Botswana-diamante gedurende die Laat Oligoseen tot Vroeg Mioseen verhinder het nie. Die resultate van die sirkoon-geokronologie het die rol van die Oranjerivier in die afplatting van die subkontinent bevestig. Die volgende model tree uit bogenoemde waarnemings na vore: diamante wat in die noordooste van die subkontinent uit kimberliete met „n voor-Karoo inplasingsouderdom vrygestel is, is aanvanklik suidweswaarts vervoer deur voor-Karoo riviere. Daarna is die diamante deur gletsers en ysplate gedurende die Dwyka-tydperk, en uiteindelik ná vrystelling vanuit ontblote glasiale en paleo-fluviale afsettings tesame met diamante wat intussen vanuit Jura- en Krytouderom kimberliete vrygestel is, deur die dreineringstelsels in die Kryt-tydperk en later, verder suidweswaarts vervoer. Sommige het onderweg in fluviale sedimente (terrasse, karstholtes en ander reste van paleokanale) agtergebly, terwyl „n beduidende hoeveelheid tot in die Atlantiese Oseaan vervoer is waar hulle deur mariene prosesse in ou strandlyne en bodemrots opvangstrukture gekonsentreer is. Die studie het ook die mikroskopiese ondersoek van 'n pakkie diamante afkomstig vanuit die enigmatiese afsettings aan die noordelike Skedelkus van Namibië ingesluit. Op grond van geologiese en geomorfologiese getuienis word die afleiding gemaak dat die Skedelkusdiamante nie met die Oranjemund-afsettings verbind kan word nie, terwyl die mikroskopiese oppervlakteksture toon dat bronne met 'n voor-Karoo inplasingsouderdom die mees waarskynlike provenans vir hierdie diamante is. Die kombinasie van FTIR-analises en oppervlaktekstuur-studies van diamante, sirkoongeokronologie en SEM-analises van granate het die formulering van „n hersiene model vir die subkontinent se dreineringsgeskiedenis sedert die Middel-Kryttydperk en diamantverspreiding moontlik gemaak terwyl die studie van sedimentêre klaste getoon het dat hoë-energietoestande, waardeur diamante van swak gehalte vernietig sou word, herhaaldelik voorgekom het, veral in die paleo-Oranjerivier. Die afleiding word gemaak dat hierdie aspek „n bydrae gelewer het tot die hoë persentasie juweelstene in die oorblywende alluviale diamantpopulasies.