Trace element fractionation trends of tholeiitic magma at moderate pressure: Evidence from an Al-spinel ultramafic-mafic inclusion suite

Trace element data on an Al-spinel ultramafic-mafic inclusion suite in an analcimite support earlier proposals that the various inclusions are comagmatic and represent fragments of a layered tholeiitic ‘pluton’ which differentiated at pressures close to 8 kb. Ultramafic inclusions are dominantly pyroxenites, often websteritic, whereas the mafic inclusions are largely two pyroxene-plagioclase assemblages. Appropriate experimental data and abundances of Sc, Cr and V indicate that subcalcic clinopyroxene or relatively Ca-rich clinopyroxene was the major ferromagnesian phase fractionated, often accompanied by spinel in the early and middle stages of differentiation and, in the later stages, by titanomagnetite. Comparatively moderate decreases in Ni and Co suggest that olivine was a relatively unimportant fractionating phase. Clinopyroxene fractionation at moderate pressures should be assigned only a comparatively insignificant role in the production of evolved basaltic compositions.     

Trace element fractionation trends of tholeiitic magma at moderate pressure: Evidence from an Al-spinel ultramafic-mafic inclusion suite

Trace element data on an Al-spinel ultramafic-mafic inclusion suite in an analcimite support earlier proposals that the various inclusions are comagmatic and represent fragments of a layered tholeiitic ‘pluton’ which differentiated at pressures close to 8 kb. Ultramafic inclusions are dominantly pyroxenites, often websteritic, whereas the mafic inclusions are largely two pyroxene-plagioclase assemblages. Appropriate experimental data and abundances of Sc, Cr and V indicate that subcalcic clinopyroxene or relatively Ca-rich clinopyroxene was the major ferromagnesian phase fractionated, often accompanied by spinel in the early and middle stages of differentiation and, in the later stages, by titanomagnetite. Comparatively moderate decreases in Ni and Co suggest that olivine was a relatively unimportant fractionating phase. Clinopyroxene fractionation at moderate pressures should be assigned only a comparatively insignificant role in the production of evolved basaltic compositions.     

Thermal and petrological structure of the lithosphere beneath Hannuoba, Sino-Korean Craton, China: evidence from xenoliths

Deep-seated xenoliths entrained in the Hannuoba basalts of the northern Sino-Korean Craton include mafic and felsic granulites, mantle wall-rock from spinel– and garnet–spinel peridotite facies, and basaltic crystallisation products from the spinel-pyroxenite and garnet-pyroxenite stability fields. The mineral compositions of the xenoliths have been used to estimate temperatures and, where possible, pressures of equilibration, and to construct a geothermal framework to interpret the upper mantle and lower crustal rock-type sequences for the region. The xenolith-derived paleogeotherm is constrained in the depth interval of 45–65 km and like others from areas of young basalt magmatism, is elevated and strongly convex toward the temperature axis. Two-pyroxene granulites give the lowest temperatures and garnet pyroxenites the highest, while the spinel lherzolites fall between these two groups. The present-day Moho beneath the Hannuoba area is defined at 42 km by seismic data, and coincides with the deepest occurrence of granulite. Above this boundary, there is a lower crust–upper mantle transition zone about 10-km thick, in which spinel lherzolites and mafic granulites (with variable plagioclase contents) are intermixed. It is inferred that this underplating has resulted in a lowering of the original pre-Cenozoic Moho (then coinciding with the crust–mantle boundary, CMB) from about 30 km to its present-day position and was due to intrusions of basaltic magmas that displaced peridotite mantle wall-rock and equilibrated to mafic granulites. Trace element patterns of the diopsides (analysed by laser ablation-ICPMS) from the Cr-diopside series spinel lherzolites and associated layered xenoliths (spinel lherzolites and pyroxenites) indicate a fertile uppermost mantle with moderate depletion by low degrees of partial melting and little evidence of metasomatic activity. The similarity in major and trace element compositions of the minerals in both rock types suggests that the layered ultramafic xenoliths formed by mantle deformation processes (metamorphic segregation), rather than by melt veining or metasomatism.     

Granulite and pyroxenite xenoliths from the Deccan Trap: insight into the nature and composition of the lower lithosphere beneath cratonic India

Granulite and pyroxenite xenoliths in lamprophyre dykes intruded during the waning stage of Deccan Trap volcanism are derived from the lower crust beneath the Dharwar craton of Western India. The xenolith suite consists of plagioclase-poor mafic granulites (55% of the total volume of xenoliths), plagioclase-rich felsic granulites (25%), and ultramafic pyroxenites and websterites (20%) with subordinate wehrlites. Rare spinel peridotite xenoliths are also present, representing mantle lithosphere. The high Mg #, low SiO₂/Al₂O₃ and low Nb/La (<1) ratios suggest that the protoliths of the mafic granulites broadly represent cumulates of sub-alkaline magmas. All of the granulites are peraluminous and light rare-earth element-enriched. The felsic granulites may have resulted from anatexis of the mafic lower crustal rocks; thus, the mafic granulites are enriched in Sr whereas the felsic ones are depleted. Composite xenoliths consisting of mafic granulites traversed by veins of pyroxenite indicate intrusion of the granulitic lower crust by younger pyroxenites. Petrography and geochemistry of the latter (e.g. presence of phlogopite) indicate the metasomatised nature of the deep crust in this region.Thermobarometric estimates from phase equilibria indicate equilibration conditions between 650 and 1200 °C, 0.7-1.2 GPa suggestive of lower crustal environments. These estimates provide a spatial context for the sampled lithologies thereby placing constraints on the interpretation of geophysical data. Integration of xenolith-derived P-T results with Deep Seismic Soundings (DSS) data suggests that the pyroxenites and websterites are transitional between the lower crust and the upper mantle. A three-layer model for the crust in western India, derived from the xenoliths, is consistent with DSS data. The mafic nature of this hybrid lower crust contrasts with the felsic lower crustal composition of the south Indian granulite terrain.     

Ultramafic inclusions and high pressure megacrysts from a nephelinite sill,Nandewar Mountains,north-eastern New South Wales,and their bearing on the origin of certain ultramafic inclusions in alkaline volcanic rocks

Ultramafic inclusions and megacrysts are unusually abundant in a nephelinite sill in the Nandewar Mountains in north-eastern New South Wales. The inclusions are divisible into a Cr-diopside group and a Ti-augite group, the former being dominated by Cr-spinel Iherzolites of restricted modal composition, the latter by olivine and titaniferous Al-rich clinopyroxene assemblages which vary widely in their modal proportions. The principal megacryst species are olivine and black, titaniferous Al-rich clinopyroxene; additional but comparatively rare megacrysts include titanphlogopite, kaersutitic amphibole, and deep green, relatively Fe-rich clinopyroxene. The Cr-spinel Iherzolites conform closely in mineralogy and chemistry with the spinel lherzolites which dominate upper mantle xenolith assemblages in alkaline mafic volcanic rocks from other provinces. Megacrysts and Ti-augite inclusion mineral assemblages are consistently more Fe-rich than analogous phases in the Cr-diopside xenoliths and also display more extensive cryptic variation. The available experimental data on the high pressure liquidus or near-liquidus phases in olivine nephelinite and related compositions indicate that the olivine and black clinopyroxene megacrysts were precipitated at pressures in the vicinity of 15–20 kb. The similarity in the nature and compositions of the principal megacryst species to analogous phases in the Tiaugite group of inclusions indicates that the latter also represent cognate cumulates derived from the olivine nephelinite at broadly comparable pressures. High pressure fractionation of the host olivine nephelinite liquid, controlled mainly by the separation of olivine and aluminous clinopyroxene, produced only comparatively minor compositional changes in the derivative liquid. The hiatus in olivine compositions at approximately Fo_86–88, apparently characteristic of the olivines in coexisting Cr-diopside and Ti-augite inclusions, is assessed in terms of the compositions of olivine in equilibrium with alkali basaltic liquids at high pressures.     

Geochemistry of mafic and ultramafic xenoliths from Fidra (Southern Uplands, Scotland): implications for lithospheric processes in Permo-Carboniferous times

Several types of xenoliths occur in a Permian basanite sill in Fidra, eastern central Scotland. One group consists of spinel lherzolites, which have geochemical and isotopic characteristics similar to those of lithospheric upper mantle from elsewhere in western Europe, with both LREE-depleted and LREE-enriched compositions. A separate group comprises pyroxenites and wehrlites, some of which contain plagioclase; these have compositions and textures that indicate that they are cumulates from mafic magmas. In terms of Sr and Nd isotope compositions, the pyroxenites closely resemble the host basanite and most likely formed by high-pressure fractionation of Permo-Carboniferous alkaline magmas at lower crustal depths. They also have mantle-like δ¹⁸O values. A third group is composed of granulite xenoliths that vary between plagioclase-rich and clinopyroxene-rich compositions, some of which probably form a continuum with the pyroxenites and wehrlites. They are all LREE-enriched and most have positive Eu anomalies; thus, they are also mostly cumulates from mafic magmas. Many of the granulites also have Sr and Nd radiogenic isotope ratios similar to those of the host basanite, indicating that they have formed from a similar magma. However, several of the granulites show more enriched isotopic compositions, including higher δ¹⁸O values, trending towards an older crustal component. Thus, the pyroxenites and granulites are largely cogenetic and are mainly the product of a mafic underplating event that occurred during the widespread magmatism in central Scotland during Permo-Carboniferous times.     

Some subcalcic clinopyroxenites from Salt Lake Crater,Oahu, and their petrogenetic significance

Some inclusions from Salt Lake Crater are essentially single-phase subcalcic clinopyroxenites whose original clinopyroxenes, prior to extensive unmixing, were tschermakitic subcalcic varieties with compositions close to Ca₃₄Mg₅₄Fe₁₂. In addition to copious amounts of orthopyroxene, very minor garnet and spinel also were exsolved from the subcalcic clinopyroxenes.The genesis of the garnet pyroxenite suite at Salt Lake Crater has been examined in terms of three models, namely: (i) cumulates from alkali basaltic magmas; (ii) fractional fusion of basanitic garnet clinopyroxenite; and (iii) anatexis of upper mantle lherzolites. Field, mineralogical, chemical and experimental data collectively favour model (iii) and indicate that the nodules are genetically unrelated to their nephelinitic hosts. The Salt Lake garnet pyroxenites can be closely equated with the garnet pyroxenites in magmatictype layers in certain alpine-type ultramafic massifs and they are also similar to many garnet pyroxenite xenoliths in alkaline volcanics from other localities.Liquids produced by anhydrous partial melting of spinel Iherzolite at pressures of approximately 20 kb commonly have picritic chemistries. The crystallization behaviour of picritic liquids at elevated pressures ( 20 kb) indicates that the initial crystallization products may be either essentially single-phase subcalcic clinopyroxenites (with minimal high pressure fractionation) or a range of olivine-aluminous orthopyroxene-aluminous subcalcic clinopyroxene-garnet-(spinel) assemblages with variable 100 Mg/(Mg+Fe) ratios (when fractionation has been operative). All these assemblages may be subsequently modified by subsolidus exsolution and recrystallization.     

Petrological constraints on the cooling history of high-temperature garnet peridotite massifs in lower Austria

High-temperature peridotite massifs occur as lensoid bodies with high-pressure granulites in the southern Bohemian massif. In lower Austria the peridotites comprise garnet lherzolites lacking primary spinel, rare garnet and garnet-spinel harzburgites, and harzburgites containing Cr-rich primary spinel instead of garnet. These phase assemblages suggest initial high-pressure equilibration and are consistent with results from garnet-orthopyroxene geobarometry indicating equilibration at around 3–3.5 GPa. Maximum temperature estimates obtained on core compositions of coexisting minerals from the peridotites are not higher than ca. 1100 °C. In contrast, pyroxene megacryst compositions, garnet exsolution textures in the garnet pyroxenites, and results from geothermometry indicate much higher original equilibration temperatures in most of the pyroxenites (up to 1400 °C). High temperatures, modal zoning, the occasional presence of Mg-rich garnetites and chemical evidence suggest that the pyroxenites are cumulates which crystallized from low-degree melts derived from the sub-lithospheric mantle. Isothermal interpolation of the high temperatures to an upper mantle adiabat suggests that the melts were derived from a minimum depth of 180–200 km. The formation of small garnet II grains and garnet exsolution lamellae in the pyroxenites and pyroxene megacrysts may reflect isobaric cooling of the cumulates from temperatures above 1400 °C to ca. 1100–1200 °C (at 3–3.5 GPa) to approach the ambient lithospheric isotherm. This model differs from other models in which the formation of garnet II was explained by an increase in pressure during cooling in a subduction zone. Isobaric cooling was followed by near-isothermal decompression from 3–3.5 GPa to 1.5–2 GPa at 1000–1200 °C, as indicated by the increase of Al in pyroxenes near garnet. Further cooling in the spinel lherzolite stability field is indicated by spinel exsolution lamellae in pyroxenes from lherzolites. The formation of symplectites and kelyphites indicate sub-millimetre scale re-equilibration during exhumation in the course of the Carboniferous collision in the Bohemian massif. The peridotite massifs represent fragments of normal (non-cratonic) lithospheric mantle from a Paleozoic convergent plate margin. Received: 22 July 1996 / Accepted 28 February 1997     

Ultramafic Xenoliths from Bullenmerri and Gnotuk Maars, Victoria, Australia: Petrology of a Sub-Continental Crust-Mantle Transition

The basanite tuffs of Bullenmerri and Gnotuk maars, Victoria,enclose abundant xenoliths of spinel lherzolites, many of whichcontain amphibole ± apatite ± phlogopite. Thexenolith suite also includes cumulate wehrlites, spinel metapyroxenitesand garnet metapyroxenites. All xenolith types contain abundantlarge CO₂-rich fluid inclusions. Microstructural evidence forthe exsolution of spinel, orthopyroxene, garnet and rare plagioclasefrom complex clinopyroxenes suggests that all of the metapyroxeniteshave formed from clinopyroxene (± spinel ± orthopyroxene)cumulates by exsolution and recrystallization during coolingto the ambient geotherm. Pyroxene chemistry implies that a rangeof parental magma types was involved. Garnet pyroxenites showa series of reactions to successively finer-grained, lower-Pmineral assemblages, which imply a relatively slow initial upwardtransport of the xenoliths in the magma, prior to explosiveeruption. The same process has allowed crystallization of phenocrystsfrom small patches of interstitial melt within xenoliths oflherzolite, wehrlite and metapyroxenite. Critically selected P-T estimates for 16 garnet websteritesare consistent with published experimental studies of the spinel/garnetpyroxenite transition, and define a geotherm from 900 °C,11 kb to 1100 °C, 16 kb. Other published data extend thecurve down to c. 7 kb and up to 25 kb. This elevated geothermsuggests that the high regional heat flow is related to convectiveheat transfer by dike injection accompanying the vulcanism.T estimates for the lherzolites range from 850–1050 °C;comparison with the derived geotherm implies that the spinellherzolites are derived from depths of 30–55 km. Thiszone has low seismic velocities (V_p = 6.8–7.8 km/sec)and has thus previously been regarded as a thick, largely maficlower crust. The xenolith data show that this Mower crust' isdominantly ultramafic, with layers, dikes and some large bodiesof pyroxenites and mafic granulites. The anomalously low V_pmay be due to the high T, the high proportion of fluid-filledpore volume, and the magnesian composition of the lherzolites.The seismically defined Moho (V_p >8.0 km/sec) coincides withthe experimentally determined position of the spinel lherzolite-garnetlherzolite transition.     

Thermal and redox state of the subcontinental lithospheric mantle of NE Spain from thermobarometric data on mantle xenoliths

Mantle xenoliths in within-plate Cenozoic alkaline mafic lavas from NE Spain are used to assess the local subcontinental lithospheric mantle geotherm and the influence of melting and metasomatism on its oxidation state. The xenoliths are mainly anhydrous spinel lherzolites and harzburgites and gradations between, with minor pyroxenites. Most types show protogranular textures, but transitional protogranular–porphyroclastic and equigranular lherzolites also exist. Different thermometers used in the estimates provide higher subsolidus equilibrium temperatures for harzburgites (1,062 ± 29°C) than for lherzolites (972 ± 89°C), although there is overlap; the lowest temperatures correspond to porphyroclastic lherzolites, whereas pyroxenites give the highest temperatures (up to 1,257°C). Maximum pressures for subsolidus equilibrium of peridotites are at 2.0–1.8 GPa. Later they followed adiabatic decompression and harzburgites registered lower pressures (1.02 ± 0.19 GPa) than lherzolites (1.41 ± 0.27 GPa). One pyroxenite gives values consistent with the spinel lherzolite field (1.08 GPa). The shallowest barometric data are in agreement with the highest local conductive geotherms, which implies that the lithosphere–asthenosphere boundary is at 70–60 km minimum depth. Higher equilibrium temperatures for the harzburgites could be explained by the existence of mafic magma bodies or dykes at the lower crust–mantle boundary. Paleo-fO₂ conditions during partial melting as inferred from the covariation between V and MgO concentrations are mainly between QFM−1 and QFM−2 in log units. However, most thermobarometric fO₂ estimates are between QFM−1 and QFM+1, suggesting oxidation caused by later metasomatism during uplift and cooling.     

Thermal history of the upper mantle beneath a young back-arc extensional zone: ultramafic xenoliths from San Luis Potosí, Central Mexico

At the San Luis Potosí (SLP) volcanic field (Central Mexico), Quaternary basanites and tuff breccias have sampled a suite of ultramafic xenoliths, predominately spinel lherzolites, spinel-olivine websterites, spinel pyroxenites, and hornblende-rich pyroxenites. Spinel lherzolites from the La Ventura maars have protogranular to equigranular textures, those from the Santo Domingo maars are strongly sheared. Both spinel-lherzolite types show similar whole-rock major and trace-element abundances. They are fertile to slightly depleted with mineralogical and geochemical heterogeneities induced by partial melting processes. Pyroxenites with either magmatic or metamorphic textures are high-pressure cumulates. Hornblende-rich pyroxenites are genetically linked to the host basanites. Most of the protogranular spinel lherzolites contain veinlets of glass along grain boundaries. These glasses are chemically homogeneous and have trachybasaltic to trachyandesitic compositions. Mg- and Fe²⁺-partitioning between olivine and glass suggests chemical equilibrium between the melts represented by the glasses and the spinel-lherzolite mineral assemblage at about 1,000°C and 10 to 15 kbar. The melts are interpreted to be of upper mantle origin. They may have been formed by in-situ partial melting in the presence of volatiles or represent percolating melts chemically buffered by the spinel-lherzolite mineral assemblage at uppermost mantle conditions. Mineral chemistry in all rock types of the whole xenolith suite reveals distinct disequilibrium features reflecting partial re-equilibration stages towards lower temperatures estimated to be from 1,050°C to 850°C at 9 to 15 kbar. The presence of similar zoning and exsolution features mainly documented in pyroxenes along with similar maximum and minimum temperatures requires all sampled xenoliths to have undergone the same temperature regime within the upper mantle. The sheared spinel lherzolites from the Sto. Domingo field are interpreted as formerly protogranular material which was sheared during uplift and cooling. The estimated mantle temperatures are higher than those predicted by low heat-flow measurements at the SLP fild, indicating that surface heat flow has not equilibrated to elevated temperatures at depth. This strongly supports a young perturbation event beneath the SLP area and connects the onset of uplift and cooling of the SLP-mantle segment with the back-arc extensional regime of the Quaternary volcanic cycle of the Transmexican Volcanic Belt.     

How late are K-feldspar megacrysts in granites?

Various petrologists have suggested that K-feldspar megacrysts grow in granites that are extensively crystallized, even at subsolidus conditions. However, experimental evidence indicates that, though K-feldspar nucleates relatively late in the crystallization history, abundant liquid is available for development of large crystals. A great deal of evidence, involving many different factors, favours a magmatic/phenocrystic origin for K-feldspar megacrysts in granites, namely simple twinning, oscillatory zoning, euhedral plagioclase inclusions, and concentric, crystallographically controlled arrangements of inclusions. In addition, abundant evidence has been presented of (1) mechanical accumulation of K-feldspar megacrysts in granites, (2) alignment of megacrysts and megacryst concentrations in magmatic flow foliations, (3) involvement of megacrysts in zones of magma mixing in granite plutons, and (4) occurrence of megacrysts in some volcanic rocks, implying that the megacrysts were suspended in enough liquid to be moved without fracturing or plastic deformation. Detailed trace element and isotopic data also indicate that megacrysts can move between coexisting felsic and more mafic magmas. Irregular overgrowths on megacrysts are consistent with continued magmatic growth after euhedral megacrystic growth ceased, the overgrowths being impeded by simultaneously crystallizing quartz and feldspar grains.     

Petrogenesis of ultramafic and mafic xenoliths from Mesozoic basanites in southern Sweden: constraints from mineral chemistry

Jurassic basanite necks occurring at the junction of two major fault zones in Scania contain ultramafic (peridotites, pyroxenites) and mafic xenoliths, which together indicate a diversity of upper mantle and lower crustal assemblages beneath this region. The peridotites can be subdivided into lherzolites, dunites and harzburgites. Most lherzolites are porphyroclastic, containing orthopyroxene and olivine porphyroclasts. They consist of Mg-rich silicates (Mg# = Mg/(Mg + Fe_tot) × 100; 88–94) and vermicular spinel. Calculated equilibration temperatures are lower in porphyroclastic lherzolites (975–1,007°C) than in equigranular lherzolite (1,079°C), indicating an origin from different parts of the upper mantle. According to the spinel composition the lherzolites represent residues of 8–13% fractional melting. They are similar in texture, mineralogy and major element composition to mantle xenoliths from Cenozoic Central European volcanic fields. Dunitic and harzburgitic peridotites are equigranular and only slightly deformed. Silicate minerals have lower to similar Mg# (83–92) as lherzolites and lack primary spinel. Resorbed patches in dunite and harzburgite xenoliths might be the remnants of metasomatic processes that changed the upper mantle composition. Pyroxenites are coarse, undeformed and have silicate minerals with partly lower Mg# than peridotites (70–91). Pyroxenitic oxides are pleonaste spinels. According to two-pyroxene thermometry pyroxenites show a large range of equilibration temperatures (919–1,280°C). In contrast, mafic xenoliths, which are mostly layered gabbronorites with pyroxene- and plagioclase-rich layers, have a narrow range of equilibration temperatures (828–890°C). These temperature ranges, together with geochemical evidence, indicate that pyroxenites and gabbroic xenoliths represent mafic intrusions within the Fennoscandian crust.     

Kimberlites and megacrystic suite: Isotope-geochemical studies

Major and rare-earth element data on Cr-poor megacrystic suite from Yakutian kimberlites were generalized. Sr-Nd isotopes were studied in garnet, clinopyroxene, and phlogopite megacrysts as well as in garnet and clinopyroxene from deformed xenoliths. It was shown that Sr-Nd composition of these minerals is similar to that in the least altered diamondiferous kimberlites. The crystallization age of megacrystic minerals was determined by Rb-Sr isochron and Ar-Ar (for phlogopite megacrysts) methods. Obtained data indicate that crystallization of Cr-poor megacrystic suite began at the prekimberlitic stage and continued to the pipe emplacement. It was established that garnets from coarse-porphyric deformed lherzolites and megacrysts are similar in major and rare-earth element compositions and were derived from a common asthenospheric source. However, the distribution of incompatible elements and P-T estimates of crystallization cannot be explained by hypothesis of fractional crystallization of garnet megacrysts. It is suggested that megacrystic assemblage crystallized directly in asthenospheric melt. En route to the surface, this melt caused a metasomatic reworking of lithospheric mantle, on the one hand, and was enriched in Mg and Cr owing to the contamination by lithospheric material, on the other hand.     

An Early Palaeozoic HP/HT granulite–garnet peridotite association in the south Altyn Tagh, NW China: P–T history and U-Pb geochronology

High‐pressure kyanite‐bearing felsic granulites in the Bashiwake area of the south Altyn Tagh (SAT) subduction–collision complex enclose mafic granulites and garnet peridotite‐hosted sapphirine‐bearing metabasites. The predominant felsic granulites are garnet + quartz + ternary feldspar (now perthite) rocks containing kyanite, plagioclase, biotite, rutile, spinel, corundum, and minor zircon and apatite. The quartz‐bearing mafic granulites contain a peak pressure assemblage of garnet + clinopyroxene + ternary feldspar (now mesoperthite) + quartz + rutile. The sapphirine‐bearing metabasites occur as mafic layers in garnet peridotite. Petrographical data suggest a peak assemblage of garnet + clinopyroxene + kyanite + rutile. Early kyanite is inferred from a symplectite of sapphirine + corundum + plagioclase ± spinel, interpreted to have formed during decompression. Garnet peridotite contains an assemblage of garnet + olivine + orthopyroxene + clinopyroxene. Thermobarometry indicates that all rock types experienced peak P–T conditions of 18.5–27.3 kbar and 870–1050 °C. A medium–high pressure granulite facies overprint (780–820 °C, 9.5–12 kbar) is defined by the formation of secondary clinopyroxene ± orthopyroxene + plagioclase at the expense of garnet and early clinopyroxene in the mafic granulites, as well as by growth of spinel and plagioclase at the expense of garnet and kyanite in the felsic granulite. SHRIMP II zircon U‐Pb geochronology yields ages of 493 ± 7 Ma (mean of 11) from the felsic granulite, 497 ± 11 Ma (mean of 11) from sapphirine‐bearing metabasite and 501 ± 16 Ma (mean of 10) from garnet peridotite. Rounded zircon morphology, cathodoluminescence (CL) sector zoning, and inclusions of peak metamorphic minerals indicate these ages reflect HP/HT metamorphism. Similar ages determined for eclogites from the western segment of the SAT suggest that the same continental subduction/collision event may be responsible for HP metamorphism in both areas.     

Ultramafic inclusions and high-pressure xenocrysts in submarine basanitoid,equatorial Mid-Atlantic Ridge

A basanitoid dredged from near St. Paul's Rocks represents the first reported occurrence of an abyssal lava containing ultramafic xenoliths and high-pressure xenocrysts. The xenoliths include three separate populations. A suite of clinopyroxene-poor spinel-harzburgite and spinel-lherzolite inclusions have tectonite textures and highly refractory mineral compositions; these nodules probably represent residua from partial melting events unrelated to the host basanitoid. One harzburgite inclusion apparently accumulated from a tholeiitic magma and is also an accidental inclusion. Ultrabasic mylonites comprise a third xenolith group whose members are comparable to lithologies exposed on St. Paul's Rocks. A diverse xenocryst assemblage including labradoritebytownite, andesine, high-Ca pyroxene, low-Ca pyroxene, aluminous Cr-spinel, Cr-free aluminous spinel and Fe-Ti oxides, suggests at least two petro-genetic affinities. Fe-rich low-Ca pyroxene and Ca-rich plagioclase, together with the cumulate harzburgite xenolith, indicate the presence of a differentiated tholeiitic suite in the area. Most xenocrysts, including Ca-rich pyroxene, spinel and andesine have alkalic affinities consistent with crystallization at elevated pressure over a wide range in temperature. Both spinel populations are interpreted as cognate precipitates with Cr-free Al-spinels crystallizing above about 10 kb and aluminous Cr-spinels co-precipitating with plagioclase at lower pressures. Compared to the olivine-dominated, low-pressure phenocryst assemblage, extraction of the generally basaltic, aluminous clino-pyroxene-dominated assemblage at high pressure should have a minor effect on the major element differentiation of the host basanitoid magma.     

Mantle derived lherzolite nodules associated with kimberlite,carbonatite and basalt magmatism: A review

Occurrences, petrography and major element chemistry of lherzolite nodules are reviewed. Spinel-garnet stability relationships in these nodules are discussed and are shown to be controlled by chemical as well as physical (P/T) considerations. On the basis of a survey of spinel compositions it is proposed that three classes of spinel bearing lherzolite nodules should be recognised: namely Al-spinel lherzolites [spinel 100 Cr/(Cr + Al)<25], Cr-spinel lherzolites [spinel 100 Cr/(Cr + Al) 25–65] and chromite lherzolites [spinel 100 Cr/(Cr + Al)>65].All lherzolite nodules yield sub-solidus P/T equilibration estimates and are interpreted as fragments of upper mantle wall rocks incorporated during the volatile charged eruptions. Depths of derivation increase from < 60 kms for most nodules in alkali basaltic magmas to > 150 kms for some nodules in kimberlites. The fact that Al-spinel lherzolites are the most common nodule type in magmas of the alkali basalt suite whilst garnet lherzolites are dominant in kimberlites is attributed to the combined effects of a typically steeper geotherm in the mantle beneath areas of alkali basaltic volcanism and a shallower depth of origin for this type of magmatism. Al-spinel lherzolites do, however, occur in the kimberlites of the western U.S.A. and south-west Greenland in contrast to their apparent absence in the kimberlites of southern Africa and Yakutia, U.S.S.R. This suggests that the uppermost mantle beneath these latter regions (as represented by nodules of chromite lherzolites and chromite or Cr-spinel harzburgites) has a more refractory residual type composition with a higher Cr/(Cr + Al) ratio—although the evidence indicates an overall decrease in the level of depletion in ‘basaltic’ magma yielding constituents with depth. Lherzolite nodules generally have chemical compositions which are depleted in such constituents relative to the pyrolite model compositions for primitive or pristine mantle; nevertheless their composition range is thought to encompass both highly ‘depleted’ and essentially ‘undepleted’ upper mantle compositions.The fact that consistent temperature estimates can be obtained from the various calibrated element exchange reaction geothermometers for some lherzolite nodules but not for others (notably those with strikingly porphyroclastic textures) may indicate that some lherzolite nodules comprise mantle derived rocks which failed to totally re-equilibrate following the diapiric movements which immediately preceded their incorporation and rapid transportation to the surface.     

Petrology of ultrabasic inclusions from basalts of Minusa and Transbaikalian regions (Siberia,USSR)

The paper presents a detailed study of the ultrabasic nodules in alkali basalts of the Minusa and Transbaikalian regions (Siberia, USSR), represented by spinel lherzolites, websterites, clinopyroxenites, plagioclase-clinopyroxene rocks, as well as fragments of large augite crystals and sanidine crystals. In the course of this study 8 new chemical analyses of inclusions and 23 chemical analyses of minerals from them have been performed. The results of the investigations imply that lherzolites and websterites are xenoliths from the upper mantle, fragments of large augite and sanidine crystals are high-pressure phenocrysts, while clinopyroxenites are cognate nodules.In the process of transportation to the surface ultrabasic nodules are subjected to strong corrosion by the basalt magma. It is noteworthy that pyroxenes and spinel are least resistant to this aggressive action, while olivine is much more stable. Because of this fact, pyroxenites have less chance of remaining intact and being found by investigators than dunites and peridotites. Therefore it is quite probable that in the upper mantle, besides spinel and garnet lherzolites, there may be present large amounts of pyroxenites, which become rather rare among the ultrabasic inclusions because they disintegrate during transportation to the surface.     

Gabbro-derived granulites from External liguride units (northern Apennine,Italy): implications for the rifting processes in the western Tethys

In Santonian-Early Campanian sedimentary melanges of the External Liguride units (northern Apennine), slide blocks of subcontinental mantle and MOR basalts are associated with lithologies derived from the continental crust. One of these sedimentary melanges, the Mt. Ragola complex, is characterized by the close association of mantle ultramafic, mafic and quartzo-feldspathic granulites. Mafic granulites show a wide compositional range. They generally display a marked metamorphic layering, but undeformed rocks which preserve a gabbroic fabric are found locally. The most frequent lithologies are Al-spinel gabbronorites, generally containing minor olivine, and Fe-Ti oxidebearing gabbronorites. Troctolites, olivine gabbronorites and anorthosites were also recovered. Relics of primary textures as well as mineral and bulk-rock compositional variations indicate a comagmatic intrusive origin for the protoliths of the mafic granulites. This intrusive mafic complex underwent a subsolidus reequilibration under granulite facies conditions, at 0.6–0.9 GPa and 810–920°C, and was derived from crystallization at intermediate levels of tholeiite-derived liquids, possibly affected by crustal contamination. Its primary features are similar to those of the upper zone of the Ivrea layered complex. The gabbroic protolith for the granulites of External Liguride units were probably crystallized into the extending Adria lithosphere in relation to the initial stages of the opening of the western Tethys.     

Mg-Al Sapphirine- and Ca-Al Hibonite-bearing Granulite Xenoliths from the Chyulu Hills Volcanic Field, Kenya

Basanites of the Chyulu Hills (Kenya Rift) contain mafic Mg–Aland Ca–Al granulite xenoliths. Their protoliths are interpretedas troctolitic cumulates; however, the original mineral assemblageswere almost completely transformed by subsolidus reactions.Mg–Al granulites contain the minerals spinel, sapphirine,sillimanite, plagioclase, corundum, clinopyroxene, orthopyroxeneand garnet, whereas Ca–Al granulites are characterizedby hibonite, spinel, sapphirine, mullite, sillimanite, plagioclase,quartz, clinopyroxene, corundum, and garnet. In the Mg–Algranulites, the first generation of orthopyroxene and some spinelmay be of igneous origin. In the Ca–Al granulites, hibonite(and possibly some spinel) are the earliest, possibly igneous,minerals in the crystallization sequence. Most pyroxene, spineland corundum in Mg–Al and Ca–Al granulites formedby subsolidus reactions. The qualitative P–T path derivedfrom metamorphic reactions corresponds to subsolidus cooling,probably accompanied, or followed by, compression. Final equilibrationwas achieved at T 600–740°C and P <8 kbar, inthe stability field of sillimanite. The early coexistence ofcorundum and pyroxenes (± spinel), as well as the associationof sillimanite and sapphirine with clinopyroxene and the presenceof hibonite, makes both types of granulite rare. The Ca–Alhibonite-bearing granulites are unique. Both types enlarge thespectrum of known Ca–Al–Mg-rich granulites worldwide. KEY WORDS: granulite xenoliths; corundum; sapphirine; hibonite; Kenya Rift