ON THE GENESIS OF PRECAMBRIAN GOLD-BEARING CONGLOMERATES

Yu. P. IVENSEN, V. I. LEVIN (Yakut Branch, Siberian Division, USSR Academy of Sciences)

The origin of gold in the conglomerates of the Witwatersrand is considered controversial. Some researchers (mainly Americans) regard the gold, uraninite, and pyrite, which are very abundant in the cement of the conglomerates, as hydrothermal minerals [13, 14, 15]. African geologists view the Witwatersrand deposits as sedimentary, essentially clastogenic, and only slightly altered by subsequent metamorphism [1, 9]. Some geologists, while accepting the sedimentary genesis of the gold, consider it not a clastogenic but a diagenetic mineral [18]. There are also compromise concepts between these extreme points of view, on which it is possible not to dwell.

Hydrothermalists, as a rule, ignore the general geological features of the distribution of gold and concentrate attention on its mineralogical characteristics: the peculiar morphology of the gold particles (films, dendrites, crystals, rare rounded grains), the presence of intergrowths of gold with pyrite, uraninite, and other minerals. They believe that the widespread occurrence of pyrite in the cement (locally up to 40% of the rock volume) and the absence of magnetite, as well as of titanium–zirconium minerals, cannot be determined by sedimentation and epigenesis and can be explained only by hydrothermal metamorphism with the introduction of sulfur. In essence, all arguments in favor of the hydrothermal origin of the gold are limited to these considerations [13, 14, 15, 16].

The sedimentary origin of the gold can be substantiated by a number of facts, chiefly geological ones: a) in the thick sequence of deposits of the Dominion, Witwatersrand, Ventersdorp, and Transvaal systems (more than 10,000 m), only a few conglomerate beds are gold-bearing, while the sequence as a whole has a gentle attitude; b) gold is distributed mainly at the base of the conglomerate beds, and the gold-enriched areas form streaks coinciding with the directions of paleocurrents; c) in the upper half of the sequence (the Ventersdorp and Transvaal systems), only basal conglomerates, transgressively overlying the rocks of the Dominion–Witwatersrand systems, are gold-bearing, and they are gold-bearing only at those points where they rest directly on the gold-bearing horizons of the underlying deposits. Consequently, the clastogenic origin of gold in the conglomerates of the two upper systems (the Ventersdorp-contact and Black Reef horizons) cannot arouse-

raise doubts; meanwhile, the morphology of the gold and its relationships with accompanying minerals here are the same as in the gold-bearing conglomerates of the underlying systems; (d) statistical processing of hundreds of thousands of samples shows the presence of a clear correlation in the distribution of gold and unquestionably clastic minerals (chromite, osmiridium), and the absence of correlation in the distribution of gold, pyrite, and uraninite; (e) the sedimentary sequence containing the gold-bearing conglomerates is cut by quartz veins. These veins as a whole are barren, but in limited intervals along their rise, where they intersect beds of gold-bearing conglomerate, gold appears in the vein quartz. Consequently, the solution from which the quartz was deposited was not originally gold-bearing; on the contrary, it borrowed gold from the conglomerate beds and did not carry it any significant distance along the rise of the vein [1, 9, 16, 19, 20, 22, 23].

The chemical inertness of gold has determined its predominantly mechanical migration in the sedimentary mantle. Coarse placer gold constitutes a very small fraction of the total amount of clastogenic gold dispersed in sedimentary rocks. Gradually abraded, it passes into finely dispersed mechanical suspensions, then into sols transported by surface waters over considerable distances, and finally into true solutions. Chemogenic concentration of gold in the sediments of terminal drainage basins undoubtedly occurs, but under very restricted facies conditions. One can speak with confidence only of chemogenic gold in the “black shale” facies of hydrogen sulfide contamination [3].

A facies series of deposits containing gold of varying degrees of dispersion can be constructed. At the beginning of the series there will be deluvium and coarse-pebble river alluvium (commercial placers with coarse clastogenic gold), and at the end of the series—“black shales” with chemogenic gold. The characteristics of gold in the intermediate links of the series (deposits of alluvial plains, river deltas, coastal facies) have scarcely been studied, but facts of the presence of clastogenic gold in fine-grained sediments are known [11].

The richest placers form in channel alluvium, under conditions of the bedrock dynamic phase. Coarse gold is in hydrodynamic equilibrium with the coarsest fraction of the alluvium. Under conditions of equilibrium and cover dynamic phases, above-bedrock, including bar, placers form, with lower concentrations of finer gold. The ratio of bedrock to above-bedrock concentrations of gold (placers) changes with movement down the valley; under the conditions of an alluvial plain, above-bedrock concentrations predominate. Still finer gold is carried into the deltaic and coastal area. According to hydrodynamic calculations, in the zone of river deltas gold particles 0.003 mm in size can no longer move by saltation along the bottom, still less in suspension in the water flow. Nevertheless, even under these conditions hydrodynamic equilibrium is manifested quite distinctly, and gold particles settle together with the coarsest fractions of the alluvium [5].

Recently, the gold-bearing Middle Proterozoic conglomerates of Karelia have been well studied. In them the conglomerates of the deltaic facies proved to be the most gold-bearing [10]. On the Vitvatersrand, the persistent horizons of conglomerates are the base of transgressive packages of rocks deposited after local or regional erosional breaks. In the rear of the transgression these conglomerates are coastal-marine, and in the frontal part—deltaic. The maximum gold content is confined to the deltas of four paleorivers [18].

The conglomerates contain such fine gold that no Quaternary placers are associated with them.

Quaternary placers are clearly associated with conglomerates of channel and piedmont-fan facies which also contain, in their bulk, very fine gold (the Neogene of southwestern Tajikistan, the Jurassic of the western Baikal region). This fact proves that, in addition to the predominant fine gold, they also contain large gold particles concentrated in Quaternary alluvium. In the Permian conglomerates of the Caucasus, or in the conglomerates of the Motskaya Formation of the Prisayan region, besides large gold particles there is also dispersed gold, which accounts for more than 50% of the total content.

The geological features of the distribution of gold on the Witwatersrand prove that we are dealing with clastogenic dispersed gold. Subsequent processes (epigenesis, early metamorphism) did not influence the general pattern of its distribution. But the morphology of the gold particles shows that the process of redeposition of gold of the type of collecting recrystallization was widespread. It requires neither a significant increase in temperature nor the action of hydrothermal solutions, since this process occurs even in some young placers [2, 8]. In the conglomerates of the Motskaya Formation of the Prisayan region and of the Mukunska Formation of the Olenek uplift, it can be seen that crystalline or dendrite-like gold particles arose as a result of the growth of gold on a rounded gold particle.

The scale of local redistribution of gold increases in those areas of the Witwatersrand where post-Ventersdorp minor folding is developed. There, industrial gold contents within the conglomerate bed are confined to the crests of anticlines, while the synclines are almost devoid of gold [17]. Uraninite on the Witwatersrand was for a long time regarded as an undoubtedly hydrothermal mineral; however, more detailed study has made it possible to regard the bulk of the uraninite, both on the Witwatersrand and at Blind River, as clastogenic [21, 22].

F. P. Krendelev [6] proposed a hypothesis concerning the existence of two genetic types of mineralization in “ancient metalliferous conglomerates”: a pyritic, hydrothermal-metamorphic type (Witwatersrand, Jacobina, Blind River) and a magnetite, clastogenic type (Tarkwa and others). To the magnetite type F. P. Krendelev assigns virtually the most diverse ancient placers in terms of composition and origin, with the exception of the three above-mentioned largest gold and uranium deposits. Therefore the boundaries of the “magnetite type” are extremely indefinite. At the same time F. P. Krendelev asserts that the “magnetite type” does not yield large gold and uranium deposits, whose presence is specific to the pyritic type and is due to the hydrothermal-metamorphic genesis of the latter.

Has F. P. Krendelev found any new arguments to substantiate the hydrothermal genesis of gold and uraninite in the conglomerates of the Witwatersrand and Blind River? F. P. Krendelev does not attempt to substantiate their hydrothermal genesis logically; instead, he lists a number of geological features of deposits of the “pyritic type.” These are: the oligomictic composition of the conglomerates (deposits of the “magnetite type,” according to his data, are enclosed in polymictic conglomerates), the presence in them of quartz-sericite cement with pyrite, metamorphism of the greenschist facies, the presence of jespilites in the enclosing sequence, and others. One cannot agree that these geological features exclude the possibility of a clastogenic genesis of gold and uraninite.

The oligomictic composition of conglomerates and the presence of jasperoids in the section cannot have any relation either to the level of metamorphism or to gold mineralization, if the latter is hydrothermal. One may cite examples of the presence of jasperoids in conglomerate–sandstone sequences of polymictic composition that contain, moreover, sericite, pyrite, and gold in the cement of the conglomerates. Such is the Proterozoic “pebble suite” of the western Baikal region [4]. In oligomictic conglomerates with kaolinite in the cement (products of reworking of a kaolin weathering crust), sericite may arise not only during early metamorphism but also during far-advanced epigenesis.

On the contrary, the oligomictic composition of the pebbles and the presence of kaolinite (sericite) in the cement become an important genetic and prospecting criterion for gold mineralization of sedimentary origin, since the largest deposits of gold-bearing and uranium-bearing conglomerates are confined to basal horizons of sequences formed after a major erosional break and overlying a thick, eroded weathering crust. This is directly recorded at Tarkwa and Blind River and must also have occurred in the Witwatersrand, judging from the composition of the clastic sequence.

Although gold and uranium are present in the deposits of the Witwatersrand, Jacobina, and Blind River, the presence of this pair of elements is not obligatory for oligomictic conglomerates with sericite–pyrite cement. In them, much more often only one of these elements is present. In the deposits named there is no correlation in the distribution of gold and uranium. At Blind River gold is a subordinate component, while in the Witwatersrand the ratio between gold and uranium changes sharply in different beds of conglomerate: the famous Main Reef, which yielded about 80% of all the gold mined, contains almost no uranium [12, 21, 22]. Thus, there are no real indications of the existence of a special “pyritic type” of mineralization in conglomerates, arising only at a definite level of metamorphism, with a characteristic accumulation of two elements: gold and uranium.

In grabens filled with Proterozoic deposits on the Aldan Shield (Atugei–Nuyamsky, Davangro–Khugdinsky, and others), sections of clastic sequences are observed that are fundamentally close to the Witwatersrand section. In them, deposits of the Udokan (Lower Proterozoic) and Maimakan (Middle Proterozoic) complexes are developed [5]. The Davangro–Khugdinsky graben, emplaced in a sequence of Archean gneisses, has been studied in greatest detail. Any intrusive or vein formations are absent in the Proterozoic deposits filling the graben*. These deposits are represented by a rhythmically structured terrigenous, predominantly continental sequence (shales, siltstones, quartzitic sandstones, gravelites, interbeds of oligomictic conglomerates).

The facies conditions of sediment accumulation alternated periodically in the following sequence: a) near-shore marine and deltaic deposits (Nognya quartzites); b) floodplain deposits of an alluvial plain and stream accumulations at the foot of a peneplain (Khugdinskaya suite, lower parts of the Davangra suite); c) lagoonal and lacustrine deposits (quartzitic sandstones of the upper Davangra suite); d) deposits of an alluvial plain and the foot of a high peneplain (lower parts of the Uyanskaya series); e) deltaic and near-shore marine deposits (upper parts of the Uyanskaya series).

* Within the Chitkandinsky graben, carbonate rocks of the Cambrian and coal-bearing Jurassic strata are cut by dikes of alkaline composition (Editor’s note).

The conglomerates are represented by three lithological types, characteristic of three stratigraphic levels (from bottom to top): (a) quartz conglomerates with quartz–hydromica cement; (b) oligomictic conglomerates with ferruginous–chlorite cement (the Sverdlov horizon of the Davangra Formation); quartz conglomerates with kaolinite–chlorite–hydromica cement. In terms of its gold content, the middle conglomerate horizon (the Sverdlov) is of greatest interest.

A study of the psephometry of the conglomerates of this horizon and of the morphology of the pebbles indicates that they belong to river alluvium. The flows that deposited the gravel were oriented across the strike of the graben (down the dip of the Proterozoic sequence). The distribution of gold within the sedimentary sequence is very regular: among samples taken from shales, gold was not found in a single sample; in sandstones it is contained in 4.2% of the samples; in gravelites, in 10.7% of the samples; and in conglomerates, in 28.0% of the samples. Stratigraphic control in the distribution of gold is also clearly visible: in the Konkulin Formation it is contained in 3.2% of the samples, and in the Davangra Formation, in 24.8% of the samples. The overwhelming majority of samples with gold are confined to the coarsest-pebble conglomerates, which formed under conditions of an active hydrodynamic regime.

The total radioactivity is also distributed unevenly through the sedimentary sequence; it reaches its highest level in the conglomerates, which correlates with the content in the rocks of clastogenic zircon and monazite, as well as with their total content of heavy fraction. All these features of the distribution of gold and of total radioactivity show that the sedimentary rocks of the Proterozoic sequence were not subjected to any hydrothermal alteration and that the gold in them is sedimentary (clastogenic). The absence even of initial regional metamorphism is indicated by the presence of kaolinite and hydromicas in the cement of the conglomerates. The crystal-chemical formula of the chlorite,

\[ \mathrm{Mg}_{0.89}\mathrm{Fe}_{2.1}^{2+}\mathrm{Fe}_{0.67}^{3+}\mathrm{Al}_{1.88}(\mathrm{Si}_{3.05}\mathrm{Al}_{0.95})_{4}\cdot \mathrm{O}_{10.72}(\mathrm{OH})_{7.28}, \]

its total iron content (75.5%), and its optical properties allow it to be assigned to aphrosiderite. On the diagram of K. B. Kepezhinskas [7], our chlorite falls within the field of reference chlorites of sedimentary rocks. This also shows that the alteration of the Proterozoic rocks has the character of deep epigenesis, but not of regional or hydrothermal metamorphism. In varieties of conglomerates containing kaolinite, a colorless mica also appears, close to sericite.

The question of the origin of pyrite, typical of the Witwatersrand conglomerates, is very important and has not yet been resolved. In the conglomerates of the Davangra Formation described here, iron hydroxides are widespread, whereas pyrite is established comparatively rarely. But here pyrite forms, in addition to disseminated impregnation, typical concretionary segregations, i.e., it is clearly a mineral of epigenesis. At the same time, pyrite is everywhere replaced by iron hydroxides; consequently, its original distribution may have been broader.

Thus, the gold mineralization in the oligomictic conglomerates with kaolinite and sericite of Proterozoic age on the Aldan Shield is, in origin, placer mineralization, just as in the Proterozoic conglomerates of Karelia and in all other conglomerates that have been studied in sufficient detail in the USSR. The genesis of the gold does not change depending on whether the conglomerates are oligomictic or polymictic, whether they contain pyrite and magnetite in the cement or do not contain them. Extensive material on foreign deposits of this type leads to the same conclusions.

Ore occurrences are known in conglomerates where the gold is undoubtedly superimposed, hydrothermal. In such cases it is estaـ

is manifested not only in the conglomerates, but also in the overlying younger rocks and in rocks of other composition. None of such ore occurrences, so far as we know, reaches the scale of an industrial deposit. Therefore there is no need to orient geologists toward the search in Precambrian conglomerates for superimposed, hydrothermal gold mineralization. On the contrary, orientation toward the search in conglomerates for deposits of the ancient-placer type, taking into account all their specific features, will undoubtedly lead to the discovery in the USSR of large deposits of this type, new for us.

LITERATURE

1. Du Toit. Geology of South Africa. Moscow, Foreign Literature Publishing House, 1957.

2. Ivansen, Yu. P. Experience in studying the morphology and microstructure of native gold. Transactions of the Zolotorazvedka and “Nigrizoloto” Trusts, ONTI, issue 10, 1938.

3. Ivansen, Yu. P. Deposits of the gold-bearing conglomerate type and prospects for their identification on the territory of the USSR. In: Geochemistry, Petrography, and Mineralogy of Sedimentary Formations. Moscow, Publishing House of the Academy of Sciences of the USSR, 1963.

4. Ivansen, Yu. P., Levin, V. I. Gold content of the conglomerates of the Western Baikal region. In: Gold-Bearing Conglomerates of the Southern Margin of the Siberian Platform. Moscow, Nauka Publishers, 1966.

5. Ivansen, Yu. P., Nuzhnov, S. V., Levin, V. I. Formational types of ancient gold-bearing placers and methods of prospecting for them. Moscow, Nauka Publishers, 1969.

6. Krendelev, F. P. Classification of ancient metal-bearing conglomerates. In: Questions of the Geology of Gold Deposits in Siberia. Proceedings of TPI, vol. 134, 1968.

7. Kenniscas, K. B. Statistical analysis of chlorites and their paragenetic types. Moscow, Nauka Publishers, 1965.

8. Kuznetsova, L. T., Simonov, Yu. G. Morphological features of placer gold of the Tungir-Amazar interfluve. Bulletin of Scientific Information of the Transbaikal Department of the Geographical Society of the USSR, no. 5, 1966.

9. Liebenberg, W. Conditions of occurrence and theory of origin of uranium minerals and gold in the ores of the Witwatersrand. In: Geology of Atomic Raw Materials. Moscow, Atomizdat, 1959.

10. Negrutsa, V. Z., Negrutsa, T. F. Lithology and conditions of formation of the Proterozoic quartz conglomerates of Karelia in connection with their gold content. Lithology and Mineral Resources, 1965, no. 6.

11. Pokrovsky, E. V. Gold content of Middle–Upper Cambrian deposits of the southwestern part of the Siberian Platform. Geology and Geophysics, 1968, no. 11.

12. Heinrich, E. W. Mineralogy and geology of radioactive raw materials. Moscow, Foreign Literature Publishing House, 1962.

13. Davidson, C. F. The present state of the Witwatersrand Controversy. Mining Mag., v. 108, nos. 2–4, 1960.

14. Davidson, C. F. Uniformitarianism and genesis. Min. Mag., v. 110, no. 3, 1964.

15. Davidson, C. F. The mode of origin of banket orebodies. Bull. Inst. Mining and Metall., no. 704, 1965.

16. Discussion on the papers: “The mode of origin of banket orebodies” by C. F. Davidson. Transactions, v. 74, no. 10, 1964–1965.

17. Greenberg, R. Author’s reply to discussion on the paper: “Structure-gold relationships as revealed by some Witwatersrand mine survey plans.” Bull. Inst. Mining and Met., no. 694, 1964.

18. Haughton, S. H. Geology and the gold basin of the Transvaal and Orange free state. Scient. S. Africa, v. 1, no. 3, 1964.

19. Ramdohr, P. Neue Beobachtungen an Erzen des Witwatersrands in Südafrika und ihre genetische Bedeutung. Abhandl. der Deutsch. Akad. der Wis. zu Berlin, no. 2, 5, 1954.

20. Reinecke, Z. The Lokation of Rayble Ore-Bodies in the Goldbearing Reefs of the Witwatersrand. Trans. and Pros. Geol. Soc. of S. Afr., v. 30, 1928.

21. Roscoe, S. M. Evidence for placer origin of Elliot Lace (Blind River), Ontario, uraniferous conglomerates. Geol. Soc. America Spec. Paper, no. 73, 1963.

22. Schilowski, M. Beiträge zur Kenntnis der radioaktiven Bestandteile der Witwatersrand Konglomerate. Neus Yharb. Mineral. Abhandl., 105, no. 2, 1966.

23. Swiegers, Y. V. Gold, Carbon, Pirite and other Sulfhides in the Black Reefs. Trans. Geol. Soc. S. Afr., v. 42, no. 35, 1940.