Carbonatite Dykes Research Articles

A rare earth element-rich carbonatite dyke at Bayan Obo, Inner Mongolia, North China

¶A carbonatite dyke, extremely enriched in rare earth elements (REE), is reported from Bayan Obo, Inner Mongolia, North China. The REE content in the dyke varies from 1 wt% to up to 20 wt%. The light REEs are enriched and highly fractionated relative to the heavy REEs, and there is no Eu anomaly. Although carbon isotope δ13C (PDB) values of the carbonatites (−7.3 to −4.7‰) are within the range of normal mantle (−5±2‰), oxygen isotope δ18O (SMOW) (11.9 to 17.7‰) ratios apparently are higher than those of the mantle (5.7±1.0‰), indicating varying degrees of exchange with hydrothermal fluids during or after magmatic crystallization. The carbonatite is the result of partial melting followed by fractional crystallization. Primary carbonatite melt was formed by less than 1% partial melting of enriched mantle, leaving a garnet-bearing residue. The melt then rose to a crustal magma chamber and underwent fractional crystallization, producing further REE enrichment. The REE and trace element distribution patterns of the carbonatites are similar to those of fine-grained dolomite marble, the ore-host rock of the Bayan Obo REE–Nb–Fe giant mineral deposit. This fact may indicate a petrogenetic link between the dykes described here and the Bayan Obo mineral deposit.

Temporal variation and carbonatite contamination in primitive ocean island volcanics from São Vicente, Cape Verde Islands

Primitive samples (MgO>5%) that are representative of the volcanic stratigraphy of São Vicente were analysed for major and trace element compositions and Nd, Sr and Pb isotopes. Mantle-normalized trace element patterns of the samples have negative anomalies of large ion lithophile elements such as Rb and K and positive anomalies for Nb, Ti and Zr, as is typical for some ocean island basalts (OIB). The oldest volcanics (OV) and the youngest volcanics (RV) are picrobasalts and basanites, whereas the intermediate age volcanics (IV) are nephelinites. Carbonatite dykes intrude the OV but are not observed in IV, indicating intrusion in the late part of the OV phase. OV have the most radiogenic Sr, least radiogenic Nd, and highly radiogenic Pb. Thus, OV comprise the most direct representation of a plume-derived HIMU component. The radiogenic isotopes of IV vary between depleted mantle (DM) and HIMU components. The isotopic compositions of RV indicate a predominance of DM. Isotopic differences between carbonatites and basaltic magmas are evidence of separate HIMU sources for these two magma types. Large negative Δ8/4 ( 208Pb/ 204Pb deviation from the Northern Hemisphere Reference Line: NHRL) is a feature of the carbonatites and basaltic rocks younger than OV. The IV also show an overall enrichment in incompatible elements, and have anomalously high CaO, P 2O 5 and Sr, and high Sr/Nd, P/Nd and P/Zr, without concomitant enrichments in Ti and Zr. The negative Δ8/4 and element enrichments of basaltic rocks younger than OV indicate assimilation of carbonatite. The assimilation of carbonatite is thought to have occurred while the basaltic magmas passed through lithosphere that was previously traversed by carbonatite.

Ba-REE fluorcarbonate minerals froma carbonatite dyke at Bayan Obo,Inner Mongolia, North China

Department of Geology, University of Leicester, United KingdomWith 3 FiguresReceived August 24, 1998;revised version accepted December 17, 1999SummaryThe new occurrence in a carbonatite dyke at Bayan Obo, Inner Mongolia, North China,of the Ba-REE fluorocarbonate minerals cebaite, hunghoite and cordylite is presented,together with their chemical composition, rare earth element (REE) patterns, andsyntaxial intergrowth relationships. These syntaxial textures, investigated by scanningand electron microprobe techniques, are interpreted in terms of the cooling historyduring the late-stage crystallization of the carbonatite magma or possible later events.The REE distribution patterns of the Ba-REE fluorocarbonate minerals in the dyke aresimilar to those of the corresponding minerals in the dolomite marble that hosts thegiant REE-Nb-Fe ore body at Bayan Obo, providing evidence that the two may becogenetic.ZusammenfassungBa-SEE Fluorkarbonat-Minerale aus einem Karbonatitgang bei Bayan Obo, InnereMongolei, Nord-ChinaDie Ba-SEE Fluorkarbonate Cebait, Huangoit und Cordylit kommen in einemKarbonatitgang bei Bayan Obo, Innere Mongolei, Nord-China, vor. Ihre chemischePresent address: School of Ocean and Earth Science, University of Southampton, UnitedKingdom

A Geochemical Study of an REE‐rich Carbonatite Dyke at Bayan Obo, Inner Mongolia, Northern China

Abstract An REE‐rich carbonatite dyke was found in Dulahala, close to the Bayan Obo superlarge REE‐Nb‐Fe mineral deposit in Inner Mongolia, northern China. The REE content in the dyke varies greatly, from 1% up to 20% (wt), which might constitute rich REE ores. Light REEs in the carbonatite are enriched and highly fractionated relative to heavy REEs and there is no Eu anomaly. The REE and trace element distribution patterns of the carbonatite are identical to those of fine‐grained dolomite marble which is the host rock of the Bayan Obo REE‐Nb‐Fe superlarge mineral deposit. This indicates a petrogenetic linkage between the REE‐rich carbonatite and the mineralizations in this region.

U–Pb zircon ages from a Devonian carbonatite dyke, Kola peninsula, Russia: a record of geological evolution from the Archaean to the Palaeozoic

Zircon from a xenolite-rich carbonatite dyke on Telyachi island, near Kandalaksha on the Kola Peninsula, northwest Russia, have been analysed for U–Th–Pb to determine the age of the dyke and study the geological evolution of the Belomorides. Zircon from the ankeritic carbonatite, analysed by conventional isotope dilution, yielded a concordant U–Pb age of 380±7 Ma. This age is confirmed by SIMS U–Pb analysis of zircon from thin carbonatite veinlets in xenoliths, which gave a concordant 238 U/ 206 Pb age of 378±8 Ma. These results demonstrate that the dyke belongs to the 360–380 Ma old Kola Alkaline Province. The xenoliths represent a variety of rock types. We have analysed zircon from xenoliths of amphibolite, gneissic granite and granite. Most of these crystals have distinct core–rim structures which are clearly shown by cathodolumenescence (CL) imaging. The majority of SIMS analysed spots in cores are close to concordant at c. 2.65–2.75 Ga, which reflects the ages of the rocks from which the xenoliths were derived. Some cores lost a large fraction of their Pb during the carbonatite magmatism. One analysed, CL-dark rim appears to reflect a high grade metamorphic event which is slightly older than 2.75 Ga. This result requires that the xenoliths also include pre-2.75 Ga rocks. Most rims developed during high grade metamorphism 1.7–1.9 Ga ago. These are CL-bright and characterized by very low Th/U ratios, and formed by a rejuvenation process in solid state which resulted in partial or total loss of radiogenic Pb. Most rims have also lost Pb during the Devonian carbonatite magmatism. The ages obtained from xenolith zircons are in excellent agreement with the independently known geological evolution of the Belomorides.

New occurrence of Ba-REE fluorcarbonate minerals at Bayan Obo, Inner Mongolia, North China, and their mineralogical features

Ba-REE fluorcarbonate minerals from a carbonatite dyke at Bayan Obo, Inner Mongolia, North China, are first reported in this contribution, in which chemical composition, rare earth element (REE) patterns, and intergrowth relationships for these minerals have been investigated. Syntactic intergrowth or syntaxy between cebaite and cordylite, as well as cordylite and huanghoite were observed. This syntactic texture resulted from the variation of chemical composition of crystallizing agents for those minerals that crystallized directly from carbonatite magmas. It is worth noting that REE patterns of the Ba-REE fluorcarbonate minerals in the dyke are similar to those of the corresponding minerals from the ore hosted dolomite marble of the Bayan Obo giant REE-Nb-Fe mineral deposit, which implies their relation in origin.

The Blue Hills Intrusive Complex in Southern Namibia—relationships between carbonatites and monticellite picrites

Geochemical and petrological investigations were carried out on ultrabasic silicate and carbonatite magmatic rocks from the Blue Hills Intrusive Complex (southern Namibia). From detailed field mapping and geomagnetic traverses we propose a model for the complex build-up of this hypabyssal laccolith. The bulk comprises mica–olivine–carbonate picrite, with numerous small offshoots into the surrounding Nama shales which are metamorphosed at the contacts to the intrusive rock. The mica–olivine–carbonate picrite is intruded by small lenses and massive layers of monticellite picrite. Locally, a phlogopite–carbonate picrite occurs, that cuts through both the monticellite picrite and the mica–olivine–carbonate picrite. Late carbonatite dykes and sills and a carbonate–phlogopite–apatite pegmatite represent the last stage of the Blue Hills magmatism. Phlogopite–carbonate picrites are regarded as the parental and nearly primary magmas. They may be formed by partial melting in the dolomite stability field at a depth of 100 to 110 km and at a temperature of 1100–1250°C. Fractional crystallisation of mostly olivine led to formation of the volatile-rich, mica–olivine–carbonate picrites, which were the first to intrude into the present position of the Blue Hills Intrusive Complex. Part of the phlogopite–carbonate picrite magma intruded undifferentiated, while part of it suffered devolatilisation and oxidation on its way to the surface which led to formation of the monticellite picrites. Both magma types intruded the hypabyssal mica–olivine–carbonate picrite. The late carbonatites most probably are residual liquids, i.e. the result of almost complete solidification of magmas like phlogopite–carbonate picrites. The carbonatitic liquids were removed from the crystal mush by filter pressing. Pegmatitic veins are the late stage residual fluids or melts from within the Blue Hills body.

Geochemical and stable isotope patterns in altered volcaniclastic and intrusive rocks of the Kruidfontein carbonatite complex, South Africa

The inner zone of the Kruidfontein Complex consists largely of carbonatitic volcaniclastic breccia with interbedded, massive to well-bedded, tuffaceous sedimentary units dipping concentrically at > 50° around the margin, the dip becoming shallower towards the centre. A geochemical study has revealed the formation of siderite, ankerite, fluorite, K-feldspar, hematite, quartz, apatite, calcite and chlorite as alteration minerals. Different generations of fluorite have been distinguished by variation in Rare Earth Element (REE) contents and Light REE (LREE) to Heavy REE (HREE) ratios. One or more of these phases correlates positively with the rare-earth minerals, whilst others show a negative correlation. The C and O isotope patterns point to three major stages of development for the inner zone of the complex. Stage one involves carbonatite magma which originated from fractionation with a H 2OCO 2 vapour phase, at 700°C in a closed system with a H 2O/CO 2 molar ratio of 0.5. The second stage of the development is recorded in the intrusive carbonatitic dykes and in at least part of the carbonatitic ash flow tuffs, by a shift to higher δ 18O values without major changes in the δ 13C values. This is a consequence of a secondary alteration for which a temperature of approximately 150°C has been estimated. The isotopic characteristic of the alteration fluid could be determined as δ 18O H2O = 1.2% and δ 13C CO2 = −3.8%; a possible source being a hydrothermal fluid which introduced isotopically light C from the deep subvolcanic area and mixed with high 18O ground water that had equilibrated at about 150°C. The third stage is characterised by siderite mineralisation, with a concomitant precipitation of fluorite. Siderite appears to be secondary to the other carbonates and can thus be assumed to have formed at temperatures at or below 150°C. The secondary nature of the siderite mineralisation is re-inforced by a negative correlation between the whole rock carbonate δ 18O values and the modal siderite content. By application of siderite/water and siderite/CO 2 fractionation factors from Carothers et al. (1988), the C- and O-isotope composition of the siderite precipitating fluid, was approximated as δ 13C CO2 = −8.0% and δ 18O H2O = −1.4%. From these findings it is concluded that siderite found in the Kruidfontein Complex is likely to have been precipitated by a late-stage Fe-bearing hydrothermal fluid, depleted in 18O and 13C. Carbon was provided by a deep-seated source within the subvolcanic region and Fe was leached from rocks that the hydrothermal fluid passed through. The siderite mineralisation and concomitant fluorite formation most probably took place during progressive cooling of the Kruidfontein Complex, as can be deduced from the positive correlation between the δ 18O and δ 13C values of siderite.

Parental melts of melilitolite and origin of alkaline carbonatite: evidence from crystallised melt inclusions, Gardiner complex

Perovskite and melilite crystals from melilitolites of the ultramafic alkaline Gardiner complex (East Greenland) contain crystallised melt inclusions derived from: (1) melilitite; (2) low-alkali carbonatite; (3) natrocarbonatite. The melilitite inclusion (1) homogenisation temperature of 1060 °C is similar to liquidus temperatures of experimentally investigated natural melilitites. The compositions are peralkaline, low in MgO (ca.␣5 wt%), Ni and Cr, and they are low-pressure fractionates of more magnesian larnite-normative ultramafic lamprophyre-type melts of primary mantle origin. Low-alkali carbonatite compositions (2) homogenise at 1060–1030 °C and are compositionally similar to immiscible calcite carbonatite dykes derived from the melilitolite magma. Natrocarbonatite inclusions (3) homogenise between 1030 and 900 °C and are compositionally similar to natrocarbonatite lava from Oldoinyo Lengai. Nephelinitic to phonolitic dykes which are related to the calcite carbonatite dykes, are very Zr-rich and agpaitic (molecular Na2O + K2O/Al2O3 > 1.2) and resemble nephelinites of Oldoinyo Lengai. The petrographic, geochemical and temporal relationships indicate unmixing of carbonatite compositions (ca. 10% alkalies) from evolving melilitite melt and continued fractionation of melilitite to nephelinite. It is suggested that the natrocarbonatite compositions represent degassed supercritical high temperature fluid formed in a cooling body of strongly larnite-normative nephelinite or evolved melilitite. The Gardiner complex and similar melilitolite and carbonatite-bearing ultramafic alkaline complexes are believed to represent subvolcanic complexes formed beneath volcanoes comparable to Oldoinyo Lengai and that the suggested origin of natrocarbonatite may be applied to natrocarbonatites of Oldoinyo Lengai.

Petrogenesis of carbonate-bearing nepheline syenites and carbonatites from Southern Victoria Land, Antarctica: origin of carbon and the effects of calcite-graphite equilibrium

Chemically-evolved carbonate-bearing nepheline syenites are intruded into basement metasediments of the Koettlitz Group on Dismal and Radian Ridges in the Pipecleaner Glacier region of Southern Victoria Land, Antarctica. Whole rock XRF data from the Dismal Nepheline Syenite defines a broad trend which is consistent with the removal of a cumulate fraction of approximate composition 70% hedenbergite, 15% nepheline, 10% titanite and 5% apatite. Stable isotope, major and trace element and mineralogical characteristics of the syenites are very similar to those of cross-cutting calcite-rich dykes indicating derivation from closely-related source magmas. The general association of carbonatites and nepheline syenites with extensional environments, suggests that the Dismal and Radian Ridge nepheline syenites and carbonatite dykes indicate a period of early Paleozoic (531 Ma) rifting or intrusion into localised tensional structures in an overall compressional regime. Assimilation of marble by the syenite magmas is evidenced by abundant rafts and xenoliths within the Dismal Nepheline Syenite, however, carbon and oxygen isotopic ratios from syenite and carbonatite calcites are distinctly lighter than values from the marble country rock indicating a magmatic source. Graphite and calcite commonly occur as aggregates in the Dismal Nepheline Syenite suggesting equilibrium between these two carbon-bearing phases. Isotopic fractionation between calcite and graphite, via the equilibrium: C + O 2 = CO 2 has resulted in enrichment of the 13C isotope in calcites from the Dismal Nepheline Syenite.

Calcite and carbocernaite exsolution and cotectic textures in a Sr,REE-rich carbonatite dyke from Rajasthan, India

AbstractA carbonatite dyke from the Sarnu-Dandali alkaline complex, Rajasthan, India, contains a remarkable suite of rare earth, strontium-rich minerals with spectacular primary textures.Sr, Mn-rich calcite in the outer 5 mm of the dyke contains exsolved lamellae of carbocernaite, (Ca,Na)(Sr,Ce,Ba)(CO3)2, orientated parallel to its twin and cleavage planes. The amount of exsolved carbocernaite increases away from the dyke margin as the Sr content of the calcite increases to a maximum 13 wt.%. Sr levels as high as this in calcite have previously been recorded only in experimental work. The carbocernaite exsolution suggests that Sr-rich calcium carbonate can be a host for major amounts of REE in carbonatite magma.Separated by a sharp internal boundary, is a complex possibly cotectic intergrowth of carbocernaite and Sr-rich calcite with late Ca-rich strontianite (19 wt.% CaO). Other minerals in the dyke include baryte, pyrrhotite, alabandite, sphalerite and occasional bastnäsite-(La) and thorite. Bands of late britholite-(Ce) traverse the dyke.The host rock for the dyke is fenitized melanephelinite which is itself traversed by narrow, <1 mm, carbonatite veins beleived to predate the carbonatite dyke. Allanite, britholite-(Ce) and rare monazite-(Ce), developed at the boundary between the carbonatite dyke and the fenite, may have been produced by a reaction between the dyke and the wall rock, or may be related to the later britholite mineralisation.The textures and mineral compositions indicate primary crystallisation. They are unique amongst rare earth-rich carbonatites which are usually late-stage phenomena with signs of secondary alteration.Comparison with experimental data available for the calcite-strontianite system suggests conditions of 500°C and 2 kbar for coexisting Sr-rich calcite and Ca-rich strontianite. A smaller scale intergrowth of calcite containing only 2.9 wt.% SrO and coexisting Ca-strontianite may correspond to a further unmixing at 350°C and 2 kbar. Since no experimental data are available for a calcite-carbocernaitestrontianite system, mineral chemistries and the interpreted sequence of crystallisation have been used to construct a hypothetical phase diagram.

Carbonatite dykes at bayan Obo, inner Mongolia, China

Calcite-rich dykes occur in the thrust fold belt near the Bayan Obo rare earth element (REE) deposit. They cut a thrust inlier of granitic migmatite within folded Bayan Obo Group sediments of Proterozoic age. Cathodoluminescence, X-ray fluorescence and microprobe studies show that the rock is a calcite carbonatite with Sr-Mn-bearing calcite, magnesio-riebeckite, apatite, pyrochlore, K-feldspar and biotite. One dyke was chosen for detailed analysis. Its margin is strongly REE-mineralized with much monazite developed adjacent to zoned apatite. Secondary alteration is marked by the introduction of Fe and Mn. The adjacent migmatite is fenitized to a magnesio-riebeckite-albite rock. The sedimentary dolomite of the Bayan Obo Group is composed mainly of Mn-Sr-RE-hearing ferroan dolomite and contains bands of opaque grains, apatite, monazite, fluorite and taeniolite. Many trace element and isotope similarities between the carbonatite dyke and the sedimentary dolomite are revealed, and the evidence supports the possibility that the dolomite is a dolomitized carbonatite tuff. The Bayan Obo REE mineralization also shows geochemical similarities with the mineralization seen in the carbonatites, and a possible genetic connection is presented.

Alkaline lamprophyre and related dyke rocks in NE Transdanubia, Hungary: The Alcsutdoboz-2 (AD-2) borehole

The Alcsutdoboz-2 (AD-2) core contains 12 magmatic dykes which belong to the Late Cretaceous lamprophyric-carbonatitic association of NE Transdanubia, Hungary. Petrographically, 11 dykes can be considered alkaline lamprophyre (mainly monchiquite), and the remainder might be called carbonatite. The lamprophyre dykes are similar to both alkaline lamprophyres and ultramafic lamprophyres in major element composition, whereas the carbonatite dyke has some features that are similar to carbonatites but others that are dissimilar. Nevertheless, both of the two types of AD-2 dykes possess higher LILE content than the ultramafc lamprophyres and kimberlites, but strongly differ from average carbonatite. Based on the REE pattern, crystal fractionation (mainly of olivine) and separation of a carbonate phase from the parental lamprophyric magma are proposed for genesis of the carbonatite dyke. These characteristics and the compositional zoning of clinopyroxene and mica suggest a complex crystallization history for these dykes. The likeliest origin for the parental lamprophyric melt is through a very small degree of partial melting of metasomatized mantle.

The Dolodau dykes, Canada: An example of an archean carbonatite

The Archean Dolodau carbonatite dykes occur near a late tectonic syenite stock located in the Northern Volcanic Zone of the Abitibi greenstone belt. The biotite sovite and amphibole-biotite silicocarbonatite dykes produced fenitization of the host rocks. The Dolodau carbonatites compare favourably with Phanerozoic carbonatites in petrography, mineralogy and geochemistry. An Archean age is suggested by field geology and the isotope data available. The similarities suggest that Archean carbonatite petrogenetic processes were similar to modern day processes.

Magma mixing and metasomatic reaction in silicate-carbonate liquids at the Kruidfontein carbonatitic volcanic complex, Transvaal

AbstractThe Kruidfontein volcanic complex is a Proterozoic collapsed carbonatitic caldera structure, the inner caldera of which is filled with carbonatitic bedded volcaniclastic rocks cut by carbonatite dykes, and the outer with bedded silicate tuffs. As well as numerous fragments of phonolitic pumice in the silicate tuffs, there are unusual banded fragments composed of alternating silicate and carbonate compositions which appear to have been originally glasses, and which give evidence for mechanical mixing of magmas which may originally have been magmas separated by liquid immiscibility. The fragments have also been strongly fenitized with the introduction of K and the replacement of Al by Fe.

The Salpeterkop ring structure, Cape Province, South Africa

Salpeterkop (32°29′S, 21°51′E) near Sutherland forms part of a dome with radial fractures extending to a distance of at least 16 km. Four concentric zones of decreasing megascopic deformation are distinguished, including a central crater 1 km in diameter, but no unequivocal shock features could be found. Intense brecciation is indicative of explosive eruptions. A radial and concentric fracture pattern has been exploited by trachytic and carbonatitic dykes. The regular outward dip of the surrounding strata is ascribed to uplift by an intrusive plug represented by numerous clast-rich trachytic apophyses at surface. The Salpeterkop structure is interpreted as a dome overlying an alkaline-carbonatite ring complex, obscured by a volcanic superstructure that has been only partly eroded.

Alkaline mafic and ultramafic lamprophyres from the Aillik Bay area, Labrador

The Aillik Bay lamprophyric dyke swarm comprises abundant sannaites, plus rarer olivine sannaites, aillikites and carbonatites. Sannaites are characterized by phenocrysts of Ti–Al titansalite plus rarer olivine and phiogopite in a groundmass dominated by pyroxene, biotite, titanomagnetite, and K-feldspar. Minor mineral phases may include apatite, nepheline, analcime, carbonate, rutile, and pyrite. Most sannaites contain well-developed leucocratic ocelli, which are commonly zoned. Olivine sannaites are similar, but olivine (Fo77–85) is more abundant, and ocelli are rarer. Both types occur together as banded dykes, and calculations support the suggestion of a common parental melt.Aillikites have zoned phenocrysts of olivine (Fo71–87) and mica in a groundmass of carbonate, apatite, mica, titanomagnetite, and perovskite. Carbonatite dykes usually exhibit textures with features akin to those of the aillikites. The mineraiogical and geochemical characteristics of the aillikites are distinct from those of kimberlites.It is the intent of this paper to provide new data on these rocks. These data are used to support one of several viable models for the generation of all dyke-rock types by partial melting of LREE-enriched mantle at depths of 90–110 km. Carbonatite dykes are believed to represent the smallest melt fraction but are the latest emplaced dykes. These small melt fractions may have followed structural weaknesses created by earlier lamprophyre magmatism.

Comb-layering in carbonatite dykes

Carbonatitic dykes from the Tertiary Kaiserstuhl volcano in the Upper Rhinegraben, FRG, commonly show fine-banded layering with skeletal branching calcite crystals oriented perpendicular to the layering. These calcite comb-layers are produced from magmatic in-situ crystallization in specific conditions of supercooling and rapid cooling rates1,2. We show here that the crystallization features underline the magmatic origin of the observed textures, mineral phases and melt compositions and are a key to the physicochemical nature of the carbonatitic magma.

Systematic Sr-isotopic variation in alkaline rocks from West Greenland

Mesozoic lamprophyre and carbonatite dykes from southern West Greenland, which are believed to have evolved from a single mantle-derived magma by crystal fractionation and liquid immiscibility, have initial Sr-isotopic ratios which are negatively correlated with the increasing contents of Sr, Nb and Zr through the suite. The isotopic variation is suggested to have resulted from contamination with radiogenic crustal Sr after the formation of the various rock types of the suite.

Gardiner intrusion, an ultramafic complex at Kangerdlugssuaq, East Greenland

The Gardiner intrusion is an ultramafic complex in the inner region of Kangerdlugssuaq fjord, forming part of the East Greenland Tertiary Igneous Province. The rocks of the intrusion can be placed into three categories. (1) An older sequence is composed of dunites, pyroxenites, and a narrow marginal zone of alkali rich rocks. The rocks of the marginal zone consist of melteigite where the bordering country rock is plateau basalt, and of contaminated rocks where the country rock is the basement gneisses. The bordering gneisses have a narrow fenitized zone. (2) A strongly alkaline and undersaturated dyke sequence comprises amphibole pyroxenites, tawites, urtites, and sodalite, nepheline and alkali-feldspar syenites. (3) A younger sequence consists of uncompahgrites (intrusive melilite rich rocks) and calcite carbonatite dykes. Parts of the uncompahgrite show layered accumulations of magnetite, perovskite and apatite on the one hand, and melilite on the other. Massive aggregate concentrations of perovskite and magnetite are also recorded. The intrusion of uncompahgrite and carbonatite has resulted in intense alkali metasomatism of the rocks of the older and the dyke sequences.