The kaolinitic clay deposits on Beaconsfield, north of Grahamstown

ArticleinSouth African Journal of Science 100(11) · November 2004with 46 Reads
Cite this publication
Abstract
The Grahamstown clay deposits occur below the Grahamstown Formation silcrete, which forms a remnant of an extensive peneplain that developed on the African erosion surface during the Cretaceous–Tertiary period. This paper provides new data on the distribution of the kaolinitic clay deposits in the Beaconsfield area north of Grahamstown. These data include 23 borehole profiles through the deposits, and the chemistry and mineralogy of the clays. Relatively little information is available on this part of the peneplain. It was found that the thickness of the kaolin horizon varies considerably, but reaches 35m in places. It generally occurs under a silcrete cover, which attains a thickness of 8 m in places. Lithological logs enable direct comparison across the Beaconsfield area. The clays are developed in both the Witteberg Group shale and Dwyka Group tillite. The contact between the clay and underlying bedrock is gradational and relatively uneven. Major-element X-ray fluorescence analyses revealed that there is chemical variation, both vertically and laterally. Al2O3 content is generally near 20%, but may reach 29%. SiO2 content varies between 55 and 70+%. SiO2 contents are highest in the silcretes occurring just below soil level. Fe2O3 is high locally in the top part of the profile. K2O and Na2O are generally low, but increase towards the unweathered bedrock as the primary feldspar content increases. These compositional variations are compatible with residual concentrations of kaolinitic clays through deep weathering below the former African erosion surface.

Do you want to read the rest of this article?

Request full-text
Request Full-text Paper PDF
  • Article
    Full-text available
    The Cretaceous period, which is considered one of the most remarkable periods in Earth’s history, saw episodes of abrupt greenhouse warming and cooling. The Cretaceous was also exceptional in that it was associated with the widespread occurrence of silcrete. To demonstrate this, the present study collated records of silcrete occurrences from the Jurassic to the present and compared them with records of variation in palaeotemperature and atmospheric CO2 levels. Quartz solubility, which is one of the key factors that controls the rate of silcrete development, was also calculated over the same period. The results demonstrate that a marked increase (approximately 100%) in quartz solubility occurred during the Cretaceous. This was found to be a direct consequence of the extreme global warmth witnessed at that time. Furthermore, this study shows that silcrete occurrences are consistent with records of palaeotemperature and that silcretes have formed mostly in regions that have experienced warm climatic conditions, with no instances found in polar regions. The Poaceae family of grass plants are known to have evolved and diversified during the Cretaceous, which coincide with the period when silica was readily available. X-ray spectra and backscattered images from SEM examination of the internal structure of a modern wild grass (which belongs to the Poaceae family of grass plants) reveals that silica is an important constituent of the grass. This suggest a possible link between evolution of the Poaceae family in the Cretaceous and the high availability of silica during the Cretaceous.
  • Article
    The town of Grahamstown is known on the geological and mineralogical point of view mainly because of its kaolin deposit, which derived from the intense weathering of the Dwyka tillite of the Karoo Supergroup. The weathering is favoured by the occurrence of brittle structures and breccias of granites that contains considerable amount of feldspar. The purpose of this study is to examine the importance of structural control in the weathering process leading to the formation of kaolin, to check the petrographical data by comparing breccias found in the fresh tillite and those in the kaolin, to highlight the mineralogical composition in some samples. The methods used in this study include: a comprehensive literature review, field observations, fault and fracture measurements to produce a general orientation, microscopic study, XRD and XRF analysis. Muscovite, albite, orthoclase, plagioclase, smectite, illite and quartz are some of the minerals present; smectization and illitization precede kaolinization in the Grahamstown area from k-feldspar and feldspar by leaching of elements such as K, Na and Ca and concentration of Al that later combines with Si to produce kaolin. A fresh tillite has higher intensity in peak diffraction analysis than a less and more weathered rock at a certain angle 2 theta. It is concluded that the primary source rock that is the parent rock in the genesis of kaolin is the Dwyka tillite, this tillite comprises breccias of granite and quartzite having microfractures that contribute to the alteration of feldspathic materials into kaolin.
  • Article
    Sacred sites are important not only for their traditional, spiritual or religious significance, but may also potentially be valuable for biodiversity conservation in human transformed landscapes. Yet, there has been little consideration of sacred sites in urban areas in this respect. Consequently, to better understand the ecosystem service and conservation value of urban sacred sites, inventories of their floral communities are needed. We examined the richness, composition and structure of the trees and shrubs in 35 urban churchyards and cemeteries in the City of Saints (Grahamstown). The combined area of urban sacred sites (38.7 ha) represented 2.2% of the city area and 13.6% of the public green space area. Species richness of woody plants was high, albeit dominated by non-native species. Levels of similarity among sites were low, indicating the effects of individual management regimens. There was no relationship between age of the site and measured attributes of the vegetation, nor were there any significant differences in vegetation among different religious denominations. However, the basal area and number of woody plants was significantly related to site size. These results indicate the significant heterogeneity of urban sacred sites as green spaces within the urban matrix. The significance of this heterogeneity in providing ecosystem services to users of sacred sites and the broader urban communities requires further investigation.
  • Article
    Full-text available
    This collaborative study has brought together our various backgrounds in art history, ceramics practice, chemistry, and clay mineralogy, so as to contextualise properties and ideas about the clay bodies used by octogenarian potters Alice Gqa Nongebeza and Debora Nomathamsanqa Ntloya. Nongebeza works from her homestead at Nkonxeni village [31º 37'59.66"S, 29º 23'22.26"E] in the Tombo area, and Ntloya is based at Qhaka village [31º36'34.04"S, 29º 24'34.78"E] in the Chaguba area, these sites being within about 5 kilometres of each other on the R61 road from Mthatha towards Port St Johns. Working separately, these potters have been digging clay from their respective mining sites and making pots for approximately the past sixty years. In this paper notions about why those specific clays were chosen and why their characteristics are desirable are enframed by comparative chemical analyses thereof as well as by a broad overview that places these clays in a wider perspective.
  • Article
    Full-text available
    Clay minerals are widely distributed in Bolivia; among them, illitic clays are most common in the Altiplano where they are mined for use in the ceramic industry that has been growing in the last few years. In addition to illitic sediments, kaolinitic sediments have been recently discovered in sedimentary units in the Bolivian Altiplano. Residual ball clay occurrences in Devonian sedimentary units were studied as part of this work. Geological mapping and geophysical studies (ERT and GPR) were done for better understanding the origin of the deposits and were part of a preliminary study of the mineral potential to define the relationship with the host rock. Chemical and mineralogical techniques such as X-Ray Diffraction, Scanning Electron Microscopy and Inductively Coupled Plasma analyses were performed in samples from the studied area to verify the presence of kaolinite. Atterberg limits and behaviour of the raw material in ceramic specimens supported by chemical analyses show that this material is suitable for manufacturing tile ceramics. This study provides fundamental knowledge for deposit exploitation and future generation of an alternative source of employment for the inhabitants of Micaya.
  • Article
    The Coastal Platform is the dominant geomorphic macro-feature of the southern Cape south of the Cape Fold Belt. The southern Cape coast has served, on-shore as well as off-shore, as an important sediment sink, since the mid-Cretaceous, and into and throughout the Quaternary. An understanding of the geomorphic evolution of the Coastal Platform since the Cretaceous provides a context for the subsequent Neogene and Quaternary geomorphic palaeoenvironments of the region. Existing evidence for the evolution of the Coastal Platform is reviewed and synthesised, and new evidence supporting the polycyclic nature of the surface is introduced. The Coastal Platform was planed under hot, humid conditions during the Cretaceous, following the break-up of Gondwana. Simultaneously, the Great Escarpment was eroded inland. This resulted in a much modified African Surface which manifests itself as the Coastal Platform extending from Bot River in the west to Port Elizabeth in the east. Evidence based on detailed map analysis is presented to show that the Coastal Platform is polycyclic. Tabular remnants of the African Surface are preserved on the Coastal Platform by laterite or silcrete duricrusts overlying variable depths of saprolite. At least four discrete surfaces can be recognised in the Albertinia–Mossel Bay area, while in the Sedgefield–Knysna area the surface is bimodal. The relationship between the Coastal Platform and two important landscape components of the region, the Bredasdorp Group limestones and the Knysna coversands, is also considered. Finally, the tectonic stability of the Coastal Platform's Cretaceous surface is questioned. Evidence is drawn from zeta bays developed over half-grabens, down-faulted to the west along northwest–southeast trending faults, and from evidence of displacement in the Cape Peninsula, as well as from stripped basal surfaces of differing altitudes. This evidence demonstrates post-Cretaceous lateral and vertical displacement on the Coastal Platform.
  • Article
    Full-text available
    This annual review of X-ray fluorescence covers developments over the period 2005-2006 in instrumentation and detectors, matrix correction and spectrum analysis procedures, X-ray optics and microfluorescence, synchrotron XRF, TXRF, portable XRF and on-line applications, as assessed from the published literature. The review also covers a survey of applications, including sample preparation, geological, environmental, archaeological, forensic, biological, clinical, thin films, chemical state and speciation studies. The trend set in previous years continues with large numbers of papers published in the areas of environmental, industrial and biological/clinical applications, followed this year by a significant contribution in the area of archaeological science and cultural heritage. Notable contributions this year are in areas of topical societal interest, including the rapid throughput screening for combatorial materials. Interest continues in nanotechnological applications as well as the related micro-analytical instrumentation that permits pro. ling and mapping of samples, which was previously undertaken by non-XRF techniques.
  • Article
    The peneplain surface is one of erosion without river deposits but is capped largely by metasomatic silcrete up to 8m thick and ferricrete. Where it cuts across argillaceous formations the latter are leached to a white clay which may exceed 30m in thickness especially on the Upper Dwyka. The various remnants of the peneplain surface isolated by erosion sometimes carry later terraces on the slopes below. Precise dating of peneplanation is impossible. In places it predates the inner portion of the Coastal Peneplain and could be Early Tertiary.-from Author
  • The Birds River Intrusion – a quantitative model for Karoo central province basalt fractionation
    • H V Eales
    Eales H.V. (1990). The Birds River Intrusion – a quantitative model for Karoo central province basalt fractionation. S. Afr. J. Geol. 93, 717–728.
  • The geology, mineralogy and chemistry of the Grahamstown clay deposits
    • J Smuts
    Smuts J. (1983). The geology, mineralogy and chemistry of the Grahamstown clay deposits. M.Sc. thesis, Rhodes University, Grahamstown.
  • Genesis of South African residual kaolins from sedimentary rocks
    • R O Heckroodt
    • D Bühmann
    Heckroodt R.O. and Bühmann D. (1985). Genesis of South African residual kaolins from sedimentary rocks. In Proc. International Clay Conference, Denver, eds L.G. Schultz, H. van Olphen and F.A. Mumpton, pp. 128-134. The Clay Minerals Society, Indiana.
  • The macro-geomorphology of the Eastern Cape In The Geomorphology of the Eastern Cape
    • R R Maud
    Maud R.R. (1996). The macro-geomorphology of the Eastern Cape. In The Geomorphology of the Eastern Cape, South Africa, ed. C.A. Lewis, pp. 1-18. Grocott and Sherry Publishers, Grahamstown.
  • The geology of the Grahamstown area. Explanation of sheet 3326
    • M R Johnson
    • F G Le Roux
    Johnson M.R. and le Roux F.G. (1994). The geology of the Grahamstown area. Explanation of sheet 3326. Council for Geoscience, Pretoria.
  • The geology of an area east of Grahamstown. Explanation of sheet 136 (Grahamstown) Geolological Survey of South Africa
    • E D Mountain
    Mountain E.D. (1946). The geology of an area east of Grahamstown. Explanation of sheet 136 (Grahamstown). Geolological Survey of South Africa, Pretoria.
  • The geology and kaolinite deposits of the Beaconsfield area
    • D Macpherson
    MacPherson D. (1998). The geology and kaolinite deposits of the Beaconsfield area, Grahamstown, South Africa. Honours thesis, Geology Department, Rhodes University, Grahamstown.
  • 1:250 000 geological series; Sheet 3326 Grahamstown
    • Geoscience Council
    Council for Geoscience (1995). 1:250 000 geological series; Sheet 3326 Grahamstown. Pretoria
  • An analysis and ore reserve estimation of the Beaconsfield kaolin deposits
    • V R Mitha
    Mitha V.R. (2003). An analysis and ore reserve estimation of the Beaconsfield kaolin deposits. Honours thesis, Geology Department, Rhodes University, Grahamstown.
  • Article
    Kaolin is of widespread occurrence in South Africa. It forms as a residue of surficial weathering of feldspar-bearing rocks in a structurally controlled geological environment. Hydrothermal activity may play a role in some instances. The deposits, methods of exploitation, beneficiation processes of the crude ore, ore reserves and economic potential of the occurrences are all described in this handbook. Considerable space is devoted in the early chapters to the chemical and physical properties of the kaolin and the testing and evaluation of the material for specific manufacturing purposes. -from Author
  • The Grahamstown peneplain
    • E D Mountain
    Mountain E.D. (1980). The Grahamstown peneplain. Trans. Geol. Soc. S. Afr. 83, 47–53.
  • Article
    The Lesotho remnant contains the type succession for Karoo low-Ti basalts of central southern Africa. The 40Ar/39Ar dating indicates that the sequence was emplaced within a very short period at about 180 Ma and consists of a monotonous pile of compound basalt lava flows which lacks significant palaeosols and persistent sedimentary intercalations. We have used geochemistry to establish a stratigraphic subdivision of the lava pile. Thin units of basalt flows, the Moshesh's Ford, Golden Gate, Sani, Roma, Letele, and Wonderkop units, with diverse geochemical character and restricted geographical distribution, are present at the base of the succession. These are overlain by extensive units of compositionally more uniform basalt, the Mafika Lisiu, Maloti, Senqu and Mothae units, which build the bulk of the sequence. A single palaeomagnetic polarity reversal occurs within the lower third of the basalt succession and is consistently located within the Mafika Lisiu unit. This and the persistent and relatively uniform thickness of the stratigraphic units suggest that the pile was constructed in a uniform manner by eruption of basalt onto a generally planar surface from a widespread plexus of dykes. The stratigraphic sequence in Lesotho closely resembles that in the thinner sequence of low-Ti basalts of the Springbok Flats remnant, some 400 km to the north. A thin unit of high-Ti basalt within the upper part of the Springbok Flats sequence can be correlated with the thick high-Ti basalt suite along the rift-related Lebombo structure on the eastern margin of the Karoo province. This is the first established correlation between these two important outcrops of Karoo volcanic rocks and demonstrates that the low-Ti basalts of Lesotho and the cratonic interior are the approximate time equivalents of the lower part of the Lebombo sequence. This conclusion has important implications for models for the origin of the Karoo flood basalt province.
  • Article
    Whole-rock trace-element analyses can be used in conjunction with partition coefficients and mineral modes to calculate a liquidus-temperature equilibrium distribution of trace elements among the constituent minerals of cumulate rocks. Once the concentration of a trace element in a given mineral phase is known, then the trace-element concentration in the liquid that was in equilibrium with this mineral can be calculated if certain assumptions are made about the temperature and major-element chemistry of the liquid. The effects of a trapped melt fraction (TMF) are accounted for by: (1) including in the calculation a phase (the trapped melt) with a mineral/melt partition coefficient of 1; and (2) adjusting the rock's mode to reflect removal of the TMF. The method yields results comparable to those given by mineral separation or in situ analysis for trace elements.
  • Article
    Mineral and rock compositions of the basal zone of olivine gabbro and picrite and the central zone of gabbros of the layered Insizwa complex at Waterfall Gorge, N Transkei, South Africa, are described and compared with similar data from the Tabankulu intrusion. These data are interpreted as evidence for successive pulses of variously differentiated magma and for the fact and location of equilibration of the parent magma with sulphide. In particular, the depleted nickel content of olivine is related to the formation of ore, its grade, tonnage and even its location (in facies context only). Application of these conclusions to the Tabankulu intrusion of the complex suggests the presence of large and rich orebodies, but they are yet to be found. -G.J.N.Dept. of Earth Sciences, Open Univ., Milton Keynes, MK7 6AA, UK.
  • Article
    Full-text available
    The importance of crystallization during compaction of an 800 m section of cumulates in Olivine-bearing zones III and IV of the Stillwater complex is assessed using whole-rock and mineral compositions. Two sets of calculations provide estimates for the liquid fraction that crystallized in each sample. The first relies upon the effect that crystallized liquid has on the whole-rock Mg#, and the second is based on the proportion of compatible and incompatible trace elements contributed by the crystallized liquid. Estimates based on trace-elements vary with their bulk distribution coefficients (D) such that elements with higher Ds yield higher estimates of crystallized liquid. Comparison of estimates based on Y (average 9%) with those based on V (average = 45%) and on the whole-rock Mg# (average 15%) show all three to be substantially different. The disparity in these estimates is interpreted to be a consequence of crystallization during compaction enriching the interstitial liquid in incompatible elements that were then expelled from the crystal pile prior to incorporation into late crystallizing minerals. Knowledge of the fate of liquid included in a growing crystal pile is of fundamental importance not only to interpreting the composition of cumulates but also to understanding how liquid compositions evolve by fractional crystallization. The return of evolved liquids from the crystal pile can decouple fractionation trends of major and trace elements producing apparently anomalous enrichment in the more incompatible trace elements.
  • Article
    Detailed petrographic, modal and geochemical studies on a number of deep boreholes (exceeding 1.2km in some instances) along the southeastern margin of the Insizwa lobe of the Mount Ayliff Complex reveal the existence of a geochemical stratigraphy in the mafic intrusive rocks. Using combined geochemical and petrographic criteria new definitions are presented for the Basal Zone and Central Zone of Scholz (1936). Within the Basal Zone a coherent sequence of ultramafic rocks (plagioclase lherzolites), previously referred to as picrites, up to 350m thick, occur and are distinguished from associated olivine-rich gabbronorites by a new set of criteria which involves bulk rock composition (MgO>30%) and petrography (cumulus olivine and spinel only). This definition can be consistently applied in all borehole cores. The Basal Zone gabbronorites have variable olivine concentrations and can be distinguished from similar rocks in the overlying central zone by Sr/Zr, Sr/Al2O3 and P/Zr and by overall higher incompatible element contents. The latter implies that Basal Zone rocks are orthocumulates and have retained a higher proportion of their residual liquid than the Central Zone gabbronorites which are mesocumulates. Within the Central Zone the gabbronorite sequence can be subdivided (from the top down) on the basis of geochemistry into a Top Gabbronorite, Cu-Poor Gabbronorite, and Upper Gabbronorite. These subdivisions can be recognized in all borehole cores. In borehole INS96-02, one of the deepest boreholes, the plagioclase lherzolite layer in the Basal Zone is ‘split’ by a sequence of gabbronorites showing chilled contacts and variable effects of thermal metamorphism. Specifically within this zone are two units of gabbronorite which are thermally metamorphosed by Basal Zone gabbronorites and which have chemical characteristics of the Central Zone. This is evidence that the Insizwa lobe was emplaced from the top down with intrusion of magmas which crystallized Central Zone gabbronorites early, followed by emplacement of Basal Zone magmas. The composition of olivine (%Fo, Ni/MgO) mimics whole-rock Mg-number and Ni/MgO variation through the sequence. Specifically olivines from the Upper Gabbronorite and the Cu-poor Gabbronorite are depleted in Ni relative to MgO, whereas olivine from the other units has ‘normal’ Ni/Mg ratios. The thick Taylor’s Koppie dyke to the E of Insizwa is built of olivine gabbronorites with variable olivine contents and scattered blebs of sulphide and is a potential feeder to the Insizwa sequence. It has whole-rock geochemical signatures and Ni-depleted olivines which are very similar to the Upper Gabbronorite of the Central Zone. The wide compositional range exhibited by contact margin samples at various levels in the Basal Zone indicates that this zone was built from doleritic magma variably charged with olivine phenocrysts and which had a wide range of incompatible element ratios and Sr-isotopic compositions. The Insizwa lobe was an open magmatic system formed by multiple intrusive events of compositionally variable magmas.
  • Article
    A map of erosion surfaces, and contour plans of elevations on the more important of these, form the basis of this paper. A consistent datum is provided by the oldest of the surfaces (the African) through the diagnostic deep weathering profiles and duri-crust cappings associated with it. The Great Escarpment was formed following continental rifting. There is conclusive evidence for the co-existence of surfaces of the same age at different elevations inland of and below the Great Escarpment. The onshore evidence of erosion surfaces is correlated with recent data on offshore sedimentation, and reveals that a single cycle of erosion prevailed from the time of rifting to the early Miocene. By the end of this period a gentle pediplain (the African surface) extended across most of S Africa at elevations of 500-600m. Most erosion and scarp recession produced thick late Jurassic and Cretaceous sedimentary sequences, but shelf sedimentation declined during the Tertiary and had virtually ceased by the Oligocene. Modest renewed uplift of 150-300m in the Miocene tilted the continent slightly to the W and initiated a new (Post-African I) landscape cycle. The cycle was terminated near the end of the Pliocene and its relatively short duration resulted in imperfect planation in most areas to levels of no >100-300m below the African surface. A 2nd uplift of major proportions at the end of the Pliocene raised the E interior of the sub-continent by as much as 900m. Major monoclinal warping resulted in the SE hinterland. The ensuing Post-African II cycle is manifested chiefly in deep incision of the coastal hinterland and downcutting along major rivers of the interior. Earlier surfaces were severely deformed and dissected. The resulting sedimentation is evident mainly in the offshore deltas of major rivers. The successive uplifts to which the continent was subject produced major changes in drainage patterns. The preservation of discrete erosion surfaces with separating escarpments over considerable periods not only confirms the episodic nature of uplifts, but indicates a dominance of backwearing over downwearing in the geomorphic evolution of the sub-continent. Some thoughts are offered towards the formulation of a tectonic model which takes account of the evolutionary scheme now established for S Africa. -from Authors
  • Article
    Equilibrium melt trace element contents are calculated from Proterozoic Nain Plutonic Suite (NPS) mafic and anorthositic cumulates, and from plagioclase and orthopyroxene megacrysts. Assumed trapped melt fractions (TMF) <20% generally eliminate all minor phases in most mafic cumulate rocks, reducing them to mixtures of feldspar, pyroxene and olivine, which would represent the high-temperature cumulus assemblage. In anorthosites, TMF <15% generally reduce the mode to a feldspar-only assemblage. All model melts have trace element profiles enriched in highly incompatible elements relative to normal mid-ocean ridge basalt (NMORB); commonly with negative Nb and Th anomalies. Most mafic cumulates yield similar profiles with constant incompatible element ratios, and can be linked through fractional crystallization. High K-La subtypes probably represent crust-contaminated facies. Mafic cumulates are inferred to belong to a tholeiitic differentiation series, variably contaminated by upper and lower crustal components, and probably related to coeval tholeiitic basaltic dyke swarms and lavas in Labrador. Model melts from anorthosites and megacrysts have normalized trace element profiles with steeper slopes than those calculated from mafic cumulates, indicating that mafic cumulates and anorthosites did not crystallize from the same melts. Orthopyroxene megacrysts yield model melts that are more enriched than typical anorthositic model melts, precluding an origin from parental melts. Jotunites have lower K-Rb-Ba-Y-Yb and higher La-Ce than model residues from fractionation of anorthositic model melts, suggesting they are not cosanguineous with them, but provide reasonable fits to evolved mafic cumulate model melts. Incompatible element profiles of anorthositic model melts closely resemble those of crustal melts such as tonalites, with steep Y-Yb-Lu segments that suggest residual garnet in the source. Inversion models yield protoliths similar to depleted lower crustal granulite xenoliths with aluminous compositions, suggesting that the incompatible trace element budget of the anorthosites are derived from remobilization of the lower crust. The similarity of the highly incompatible trace elements and LILE between anorthositic and mafic cumulate model melts suggests that the basalts parental to the mafic cumulates locally assimilated considerable quantities of the same crust that yielded the anorthosites. The reaction between underplating basalt and aluminous lower crust would have forced crystallization of abundant plagioclase, and remobilization of these hybrid plagioclase-rich mushes then produced the anorthosite massifs.
  • Article
    For accurate X-ray spectrographic analysis of geological samples of widely varying composition, a fusion with lithium borate containing lanthanum oxide is used to make a suitable glass disc. Mg, Al, Si, P, K, Ca, Ti, Mn and Fe are determined on this disc, using similar discs made with pure chemicals for calibration. “Nominal percentage” composition of the sample is obtained as a direct reading from the scaler, counting time adjusted to give 104 or 103 counts for 1 per cent. For normal silicates this value will be within ±3 per cent (relative) of the correct value. For more variable samples and for greater accuracy with silicates, matrix corrections are made using the “nominal percentage” in a first order equation. The matrix correction coefficients have been determined experimentally and are shown to agree well with theoretical values. They are presented in a form suitable for use in other laboratories. For Na, a separate pressed powder sample is prepared and matrix corrections are not necessary.The precision of the method has been studied and is largely controlled by counting errors. An estimate of the accuracy has been made by comparison of X-ray results with two sets of published data on five “standard” rock samples; observed differences could all be explained as due to precision errors in the X-ray and chemical results. The coefficient of variation between the X-ray and chemical results was less than 1 per cent where the concentration of an element, expressed as an oxide, was greater than 1 per cent.Results are given to show that when matrix corrections are applied, samples with extreme compositional variability, e.g. rocks, minerals or ores, may be analysed satisfactorily.With the borate discs, background determinations are not made on each sample so that determinations involve reading the Kα line only. Rapid analyses are therefore possible; one operator can prepare and analyse more than 12 samples per day for 10 elements.