Witwatersrand Gold DepositsTHEIR GENESIS IN THE LIGHT OF MORPHOLOGICAL STUDIES

D. K. HallbauerMining Technology Laboratory, Chamber of Mines of South Africa

Studies of the morphology and other properties of the gold particles

in the fossil placer gold deposits of the Witwatersrand provide

fascinating glimpses of the conditions under which they were formed.

In particular they indicate that in some areas primitive plant life on

the deltas on which the gold was deposited acted not only like the

legendary golden fleece in concentrating fine gold particles, but

probably also reacted biochemically with these particles to determine

the forms in which gold is observed in the carbonaceous remains of this

plant life today.

In geology ancient events may be studiedthrough their effects on the natural substances ofthe Earth (1). From the disposition of thesesubstances and from their composition and mor-phology we can make certain deductions as to thenature of the events that produced them. Thisapplies not only to large bodies of rock but also tothe mineral grains of which they are composed.

In the 2 500-2 800 million year old fossil placers(2) of the Witwatersrand there lies evidente — inthe morphology of the individual gold particleswhich they contain — of the nature of the processeswhich led to their formation. Thus recent studies(3) have shown that most of the gold particles inthe reefs retain their detrital morphology, andfrom extensive studies of this morphology im-portant information on the genesis of the depositshas been derived.

The major processes influencing the distributionof gold in the conglomeraten have been establishedas(1) Alluvial transport from a primary source into

an inland basin surrounded by mountains(2) Concentration by wind and water action(3) Partial redistribution by biochemical processes

within a primitive biological environment and(4) Partial remobilisation and redeposition by geo-

logical processes during metamorphismAll these processes have affected the mor-

phology of the gold particles in Witwatersrandrocks, and three types of gold can be differentiatedin the reef, namely, detrital gold, `biochemically'deposited gold and recrystallized gold.

Detrital GoldGold occurs in its primary deposits mostly as

crystalline particles of a more or less complex,jagged morphology, sometimes intergrown withother ore minerals. During the weathering processand subsequent alluvial transport the malleablegold is increasingly deformed by collisions withthe accompanying debris. One of the extremeexamples where gold is recognised as having beentransported for more than 150 km is the famous`Rhine gold', found in the upper parts of the riverRhine in Germany. Particles recovered from gravelnear the town of Karlsruhe are examples of thisflaky, heavily deformed gold (Figure 1). No traceof the original crystalline morphology is left, and

Fig. 1 Flaky gold particles from the Rhine River, Ger-many. The structures of these particles are typical ofthose developed during extended fluvial transport

18

Fig.2 Jagged, partially flat-tened gold particles re-covered from a specimen ofB-Reef, Loraine Gold Mine,Orange Free State

the particles appear to have been flattened andbent several times during river transport. Charac-teristically, several lace-like wafers of gold may bepressed together to form such particles.

Considering the long geological history of theWitwatersrand rocks it is surprising to find thesame and similar micro-textures on the ancientgold particles (Figure 2) after they have beengently liberated from the rock by disintegratingthe siliceous matrix in hydrofluoric acid (4). Evenrare structures such as those formed by gold beingfolded around mineral inclusions can be ob-served (Figure 3).

However, most of the gold particles in theWitwatersrand deposits display micro-textureswhich indicate short transport distances of arange from 5 to 30 km (3). Depending on theoriginal shape and size of a particle the initialdistortion manifests itself in bending and hammer-ing of the protruding parts, accompanied byscratching and gouging of the surface in alldirections. The distortions and abrasions are morepronounced on larger grains and the original cry-stalline texture is deformed with increasing Bis-tance of transport, to such an extent that it maybe completely obliterated. The example shown

Fig. 3 Partial views in a scanning electron microscope of gold particles deformed around mineral inclusions. On theleft is a particle from a Transvaal river, and on the right is a particle from a sample of Basal Reef, St. Helena Gold Mine,Orange Free State

19

Fig. 4 SEM photograph of an angular goldparticle from the B-Reef, Loraine Gold Mine,Orange Free State, showing the initial stages ofabrasion and deformation during fluvial trans-port

in Figure 4 is that of an angular gold grain withremnant crystalline texture but abrasion mani-festing itself along the edges and protruding parts.By comparison with recent alluvial gold grains thedistance of transport for this and other grains fromthe same locality is estimated at about 10 km.

Gold grains under the conditions of recentalluvial transport are subjected not only to me-chanical distortion but also to chemical attack bysulphate-containing river water which normallyhas a pH value of 6-8 and an oxidation-reductionpotential of about + 300 mV. Under these condi-tions silver is leached out of the gold, thus in-creasing its fineness F = 1000 Au/(Au + Ag) withincreasing distance of transport. No such leaching

of silver from gold particles of the Witwatersranddeposits has been observed, and geochemicalconsiderations point to their having been trans-ported in a chemical environment with oxidation-reduction potential below —100 mV and pH 6-8.The fineness measured for gold particles in Wit-watersrand deposits can therefore be taken as theoriginal fineness. This is of potential interest in thegeochemical identification of reefs according to thefineness of their gold particles. This often changesboth vertically and laterally from one reef horizonto the other and indicates that the often voicedopinion that younger reefs generally represent thereworkings of older eroded reefs (5) may only betrue for isolated localities. Derivations from geo-

Fig. 5 A specimen of BasalReef from the St. HelenaGold Mine, Orange FreeState, showing a carbonlayer with trapped, fine -

grained detrital gold

20

logically related but geochemically different pri-mary deposits, or from geochemically differenthorizon in the one primary deposit, could beresponsible for the variations in fineness whichare observed.

More information on the nature of the primarydeposits has been obtained from typical detritalgold grains intergrown with other minerals. So farintergrowths between gold and sphalerite, cobal-tite-gersdorffite, chalcopyrite, bravoite, pyrite,pyrophyllite and indirectly with molybdenite havebeen observed. These mineral associations aresomewhat different from those commonly foundin the Precambrian rocks of the Barberton Moun-tain land of the Eastem Transvaal, which arecommonly taken as a provenance model for theWitwatersrand and point to a type of primarydeposit similar to porphyry-type deposits.

It is of interest in this connection that evenlome of the older sedimentary rocks of the Swazi-land sequence in the Barberton Mountain land, theMoodies series, contain small amounts of detritalgold recognizable by the morphology of theparticles. This points to the existence of sedi-mentary, although subeconomic, gold deposits ofan age greater than 3 300 million years.

`Biochemically' Deposited GoldLayers of carbonaceous matter which often

accompany the gold bearing strata on the Wit-watersrand have been recognised as fossilizedPrecambrian plant mats (5). In these mats, whichoccupied the lower parts of the alluvial fans, goldparticles were clearly trapped in the same way asthey may be trapped by corduroy tables inmetallurgical practice, as is evidenced by the manydetrital particles which can normally be recoveredfrom layers of carbonaceous matter. Especially

fine particles of gold were held back in these matsand consequently layers closely packed with fine-grained gold can occasionally be observed (Fig-ure 5).

Apart from such fine-grained gold, and com-pletely locked up within the carboxiaceous matter,however, filamentous gold and aggregates of goldfilaments occur which point to an intra- and extra-cellular concentration of gold by the primitiveorganisms constituting the mats. Although thereare no examples from more recent times of theactive participation of such plants in the redistri-bution of gold, there are indications that the goldparticles were penetrated and partially or whollydissolved by biochemical processes. X-ray radio-graphs of the carbonaceous matter show the re-lationship between detrital gold and other detritalminerals on the one hand and the biochemicallyconcentrated gold on the other. While detrital goldand other heavy minerals are concentrated in theupper portions and on top of the carbonaceouslayer, gold filaments are distributed throughoutthe fossil plants.

It may be mentioned in this connection that allcarbonaceous matter in Witwatersrand depositscontains up to 10 per cent U 30 8. Although this ismostly in the form of uraninite inclusions, finefilaments of U,O 8, which sometimes show wellpreserved cell structures, can also be observed.Such observations indicate that the primitiveplants probably brought about a re-distribution ofuranium by biochemical processes, similar to thatwhich is observed in the case of gold.

The biochemically deposited gold does notexhibit any of the features of the detrital gold. Itoccurs as irregularly shaped, often fibrous parti-cles which often have the appearance of agglutina-ted, interwoven filaments or perforated flaky par-

Fig. 6 SEM photographs of a soit fungus (A) with inorganic encrustation (Si, Al, Fe) and a gold filament (B) extractedfrom carbonaceous matter from a Witwatersrand deposit. The gold filament is a fossil gold incrustation on a hypha

21

Fig. 7 Secondary growth of gold on a pyritecrystal from a sample of Ventersdorp Con-tact Reef, Western Deep Levels Gold Mine

ticles. Most filaments have an average diameter of1-2 µm. They are often branched and occasionallyshow septation similar to algal or fungal filaments.

A comparison of these filaments with inorganicencrustations on some present-day soil fungi(Figure 6) reveals a remarkable similarity. Manymodern algae and fungi, the latter especially inlichenized form, have the ability to absorb in-organic materials, including heavy metals (7) fromthe environments in which they live, and to formextracellular deposits derived from them. A similarprocess of digesting and redistributing gold anduranium could therefore be assumed for theancient forms in the Witwatersrand rocks. Anadditional indication of a biochemical redistribu-tion of the gold is the difference in the silvercontents of detrital and `biochemical' gold.Silver is often enriched in the latter, particularlyin samples of Basal reef from the Orange FreeState, to concentrations which may be twice thosein the associated detrital gold. On the other handmercury, which is a principal trace element indetrital gold, appears to have been rejected by theorganism leaving the biochemical gold with onlysmall traces of this element.

From such geological observations it is evidentthat the most abundant plant during Witwaters-rand times, named Thuchomyces lichenoides (8),thrived in habitats such as those provided by thelower portions of wet alluvial fans, and shallowpools, and that during dry times it survived desertconditions.

Evidence for the existence and active participa-tion of living organisms in the redistribution ofheavy metals during Precambrian times is notconfined to the Witwatersrand.

The Blind River Conglomerates of the Elliot

Lake district in Canada, for example, containcarbonaceous matter enriched in uranium andthorium which show morphological features similarto the fossil plants of the Witwatersrand, togetherwith structures that can be intepreted as fossilizedbacteria.

Recrystallized GoldWhile the detrital gold and the biochemically

deposited gold are contemporary with the rocksin which they occur, the recrystallized gold is theproduct of later, i.e. younger, processes. Althoughit occurs in comparatively small amounts, itspresence is nevertheless a reflection of otherprocesses which have contributed to the genera-tion of the Witwatersrand deposits.

The mild metamorphism to which the rockshave been subjected has evidently not been strongenough to mobilize and distribute the bulk of thedetrital gold. Locally, however, near volcanicdykes and fault zones and immediately belowthick covers of Ventersdorp lava the temperaturenof the rock have clearly been sufficiently raised tomobilize and transport gold for short distances.

The resulting secondary gold occurs almostinvariably as an overgrowth on secondary pyrites(Figure 7). The pyrite which has crystal latticedimensions similar to those of gold has acted as abase for epitaxial growth of the gold. Although goldovergrowths on pyrite are rarely well crystallized,sufficient particles have been observed to indicatethat the cube faces of the gold crystals grow on thecube faces of the pyrite rotated by 45

Genesis of the Witwatersrand Gold DepositsAlthough at one time the subject of controversy,

the fossil placer nature of the Witwatersrand gold

22

Fig. 8 Artist's impression of part of thepaleo-Witwatersrand with the surroundingmountains in the background and a wetalluvial fan in the foreground. The atmos-pheres of Precambrian times were low inoxygen and transmitted blue light preferen-tially. Hence the dominant blue colour ofthis painting by Dr. F. Kaempfe, Johannes-burg

deposits is accepted without reservation by geo-logists actively involved in study of them. Themorphological and other studies of the goldgrains which they contain provide convincingsupport for this view.

The gold bearing conglomerates, or reefs as theyare commonly called, represent the final stage of adynamic series of processes which, although some-what more complicated, is nevertheless similar toothers which we observe daily. If, for example, wenotice a smashed flower pot on the pavement infront of a multi-storey building the reconstructionof the dynamic processes which led to the forma-tion of the pot fragments comes quite easily to us.To determine the original direction of flow in adry or fossilized river bed requires more detailedobservations and deductions. Many years of obser-vation by numerous geologists have been necessaryin order to enable us to reconstruct to some extentthe processes which led to the formation of thelargest known gold deposit on earth. The observa-tions on the gold morphology and its significanceare but one step in this reconstruction process.

The distribution of gold and uranium in theWitwatersrand deposits is essentially controlledby the paleo-environment. Like any other ecolo-gical system the palen-Witwatersrand consisted ofsmaller systems in which local environmentsdetermined the distribution of economic minerals.

With our present knowledge we perceive thepaleo-Witwatersrand as a basin surrounded bymountains of moderate height which were geo-logically separated from the low-lying area byactive faults. Debris washed into the basin formedalluvial fans at the base of the mountains. Thesefans extended with rapidly decreasing slope anglesfor several kilometres into the basin. Sheet floods

and a system of braided streams reduced theslope angle with time to the average value of about0.5° which is characteristic of the Carbon LeaderReef, the Vaal Reef, Basal Reef and others. Asindicated by the occurrence of wind-blastedpebbles, the so-called dreikanters, in various partsof the reefs, much of the surface of the fans wassubjected at times to desert conditions.

Wind action and water action slowly removedthe fine-grained and lighter components of theoriginal gravel deposit until the conglomeratelayer, with its accumulation of heavy mineralswhich is being mined today, remained.

At the lower portions of the alluvial fan biolo-gical activity in shallow water and stagnant poolsproduced thin mats of primitive plants whichconcentrated gold and heavy minerals bothmechanically and biochemically.

Figure 8 is an attempt by Dr. F. Kaempfe, todepict the paleo-environment of a small region ofthe Witwatersrand. Although it is an artist'screation many a geologist has contributed to itby his puinstaking and detailed observations.

References1 D. L. Eicher, `Geological Time', Foundation of Earth

Science Series, Prentice Hall Inc., Engelwood Cliffs, NewJersey, 1968

2 D. A. Pretorius, Minerals Sci. Engng., 1975, 7, (1), 18-47

3 D. K. Hallbauer and T. Utter, Miner. Deposita, 1977, (inpress)

4 G. J. Neuerburg, Jour. Research U.S. Geol. Survey, 1975, 3,(3), 377-378

5 H. C. M. Whiteside et al., in 'Mineral Resources of theRepublic of South Africa', Geological Survey, Pretoria,1976, 42

6 D. K. Hallbauer, Minerals Sci. Engng., 1975, 7, (2), 111-131

7 W. Ernst, 'Schwermetallvegetation der -Erde', Gustav Fis-cher Verl., Stuttgart, 1974

8 D. K. Hallbauer et al., Geol. Rundsch., 1977

23