Summary in English. / Includes bibliographical references. / In 1993, dredge-sampling from 40.6°S to 525.5°S, during cruise 9309 of the R/V Maurice Ewing, greatly increased the number of basalts recovered from this section of the southern Mid-Atlantic Ridge (SMAR), and significantly expanded the global suite of samples from slow-spreading ridges.

Bibliography : pages 261-277. / Rocks of the Namaqua Metamorphic Complex are exposed in an area of 10 000 km² in the northwestern part of the Namaqua Mobile Belt east of Liideritz. The Garub sequence represents the oldest rocks in the Aus area and comprises a diverse group of layered rocks of mainly semi-pelitic, pelitic, mafic, calcareous and quartzose composition. These rocks have been metamorphosed to form marbles, calcitic gneisses, metaquartzites, biotite schists, sillimanitecordierite garnet gneisses, amphibolites and granolites with minor amounts of iron formation and magnesian rocks. The principal metasediments are concentrated in west-trending zones. A central zone of calcareous rocks broadens westwards and contains the largest known bodies of carbonate rocks in the Namaqua Metamorphic Complex. The calcareous zone is bounded in the north by a narrow beit of quartzose rocks and in the south, by a broad zone of aluminous rocks. Gradational rock types between these zones are compatible with original sedimentary lithofacies changes. Layered biotite gneiss of psammitic composition has been interfolded with units of the Garub sequence. Both these rock units are present as inclusions within a tonalitic augen gneiss.

The current study focuses on the Lower Zone of the Bushveld Complex, South Africa using multiple geochemical and isotopic systems to determine the origin of crustal signatures, i.e. crustal assimilation or recycled crust in the mantle source, present throughout the Rustenburg Layered Suite (RLS) such as elevated  18O values. These geochemical and isotopic systems include major elements, trace elements, highly siderophile elements, oxygen isotopes, 87Sr/86Sr, 143Nd/144Nd, and Os-Os isotopes. Samples come from the Nooitgedacht Borehole 2 (NG2) at Union Section of the western limb of the Bushveld Complex, which sampled the Lower Zone. The 87Sr/86Sri (0.7043 – 0.7086) and Ndi values (-7.40 - -4.97) calculated in this study are in agreement with published data for the Bushveld. The majority of NG2 samples contain  18O greater than mantle peridotite (5.50‰) and MORB (5.70‰), in which NG2  18O ranges from 5.60 up to 8.00‰ for olivine, orthopyroxene, and clinopyroxene separates. These high  18O values suggests the Bushveld magmas sourced a crustal reservoir, either through crust assimilation or recycled crustal materials in the mantle source. This is the first study utilizing the Re-Os isotope system for the Lower Zone. The Osi values for the NG2 suite range from -4.37 to +35.9, which overlap with published data for the Critical Zone and the Platreef, the only previous Re-Os studies on the Bushveld. However, there are no previously reported negative Osi values for the Bushveld. The range in Osi values for the NG2 samples suggest mixing of at least two geochemical reservoirs. In addition to Lower Zone NG2 samples, sample NG2-773.65 is a chilled margin sample at the base of the NG2 borehole that contains high  18O ( 18O = 9.42 – 9.78‰) and radiogenic Osi (Osi = +62.5), in which the crust and/or recycled crust in the mantle source caused these high values. Sample NG2-734.64 contains the second lowest  18O ( 18O = 5.67‰) and most unradiogenic Osi (Osi = -4.37) values for the NG2 suite, evidence for a harzburgitic SCLM (H-SCLM) mantle source component. A lack of correlations for Osi values with  18O and 87Sr/86Sri values are indices for crustal assimilation processes. This lack in correlation may suggest a crustal component in the mantle is more likely than assimilation of the crust during ascent of the magmas toward the surface. The geochemical data presented in this study suggest the Lower Zone parent magma had a H-SCLM mantle source component in addition to the assimilation of the crust and/or the eclogitic SCLM (E-SCLM).

Presented here are major and trace element, stable (oxygen and hydrogen) and radiogenic (Sr-Nd-Pb) isotope analyses for a Cr-poor megacryst suite from the Salpeterkop complex, South Africa. The clinopyroxene, amphibole, phlogopite and ilmenite megacrysts all appear to be cogenetic, and based on known mineral relationships and intergrowths from xenoliths in the complex, the apparent order of mineral crystallisation is as follows: phlogopite → ilmenite → amphibole → clinopyroxene. Megacrysts of amphibole and phlogopite exhibit δD and δ18O values that are aligned with these grains having crystallised from melt originating from the upper mantle. Additionally, the amphibole and phlogopite megacrysts appear have experienced dehydration styled degassing, possibly related to their exhumation. Calculated P-T conditions have the megacrysts crystallising in the lower crust, under conditions ranging from 1 to 1.5 GPa (35 to 45 km depth) and 1000 to 1250 ℃. Calculated REE melts in equilibrium with the megacryst as well as radiogenic isotope results suggest that the Salpeterkop ultramafic lamprophyres are genetically related the the SPKC megacryst suite, however, the calculated parent melt to the megacryst appears to have mixed with a HIMU component. These findings primarily affect higher Mg-number megacrysts, suggesting that this assimilation or mixing occurred during initial stages of crystallisation. Lower Mg-number megacrysts lack the variations noted in their more primitive counterparts and present more tightly defined trends. A model of formation for the megacryst suite of the Salpeterkop complex sees grains having crystallised from an SPKC ultramafic lamprophyre-like melt originating from sublithospheric/asthenospheric conditions. During ascension the melt episodically assimilates material with a HIMU signature. The high Mg-number megacryst population crystallises from this melt at lower crustal depths. Soon after assimilation halts the megacryst parent melt homogenises (or re-homogenises), with grains to crystallise from this melt forming the low-Mg megacryst population.

Includes abstract. / Includes bibliographical references. / This study argues for the formation of bonebeds by aggregation of sub-adult Lystrosaurus during extended episodic periods of extreme climatic conditions, such as cold or drought, in the earliest Triassic Karoo Basin.

Includes bibliographical references. / Eighteen Jurassic to Cretaceous South African kimberlites representative of group 1, group 11 and transitional varieties that have been emplaced through both the Archean Kaapvall craton (on-craton) and Proterozoic Namaqua-Natal belt (off-craton), have been selected for a comparative study aimed at characterising their geochemistry and source region compositions, as well as understanding the petrogenetic processes that have affected them. The petrography of the analysed kimberlites is similar to typical group 1 and group 11 kimberlites, characterised by deformed and anhedrarl olivine and phlogopite macrocrysts, with more subhedral to euhedral olivine and phlogopite phenocrysts and microphenocrysts, set in a groundmass of mostly serpentine, calcite and phlogopite (group 1 kimberlites), or calcite, serpentine, phlogopite and diopside (group 11 kimberlites). The transitional kimberlites tend to show intermediate characteristics, with the on-and off-craton transitional kimberlites showing more similarity to group 1 and group 11 kimberlites, respectively.

The Andean continental arc is built upon the thickest crust on Earth, whose eruption products reflect varying degrees of crustal assimilation. Physical interactions between mafic and felsic magmas for this arc are therefore difficult to recognize due to the differentiation of mantle-derived magma during ascent through the thickened crust and a corresponding lack of erupted primitive lavas. However, a rare concentration of less evolved rocks is located marginal to the partially molten Altiplano-Puna Magma Body (APMB) in the Altiplano-Puna Volcanic Complex of northern Chile, between 21º10'S and 22º50'S. To unravel the petrogenesis of these less evolved eruptives and their spatiotemporal assessment, this work makes use of new major and trace element data, and Sr and Nd isotope ratios of fourteen Pliocene to Quaternary volcanoes. Whole-rock compositional and Sr and Nd isotope data reveal a large degree for compositional heterogeneity (e.g. SiO2 = 53.2 to 63.2 wt%, MgO = 1.74 to 6.08 wt%, Cr = 2 to 382 ppm, Sr = 304 to 885 ppm, 87Sr/86Sr = 0.7055 to 0.7088, and 143Nd/144Nd = 0.5122 to 0.5125). The least evolved products erupted along the periphery of the APMB and are likely equivalent to the replenishing magmas that thermally sustain the large APMB system. Here it is found that the investigated mafic to intermediate eruptives reflect mafic melt injections that underplate the APMB and escape along the sides of the large felsic body to avoid significant compositional modifications during ascent (e.g. La Poruña, San Pedro, Palpana, Chela volcanoes). Investigating these magmas therefore assist in assessing the evolution of the APMB through space and time. Additionally, individual volcanoes demonstrate that contamination of parental melts was caused by different differentiation processes (e.g. magma mixing, fractional crystallization, crustal assimilation) as a function of their spatiotemporal framework relative to the APMB and its melt fraction variation from peripheral (ca. 4 vol%) to central (up to 25 vol%) portions. New insight into the petrological processes that governed the magmatic evolution of La Poruña, one of the centres that exhibit the least evolved eruptions in the area, is presented. La Poruña is a 100 ka scoria cone, composed of pyroclastic material and an extensive basaltic-andesite to andesite lava flow, with magmatic evolution related to the neighbouring larger San Pedro stratovolcano. New petrography, geochemistry and radiogenic isotope data describe a set of porphyritic mafic samples, comprising olivine- and pyroxene-rich rocks, with well-defined major element compositional trends, as well as trace and rare earth element characteristics, that reflect magmatic differentiation at midupper crustal levels. Additionally, magma mixing as well as assimilation and fractional crystallization processes acted on these La Poruña magmas. A remarkable compositional feature is the unusual reversed isotopic behaviour of increasing silica with decreasing 87Sr/86Sr compositions related to later magmatic evolution involving selective assimilation during turbulent ascent at shallow crustal levels prior to eruption, therefore differing from the broadly accepted Central Andean magmatic model. In order to robustly model magma evolution and assimilation at subduction zones such as the Andes, the compositions of parental magmas feeding crustal magma reservoirs need to be defined. Here, new olivine and clinopyroxene oxygen isotope data from six volcanoes located at the margins of the giant APMB is presented, as these data provide robust constraints on parental magma genesis and further contamination processes. Existing olivine and pyroxene  18O values for the Central Andes are highly variable and potentially not representative of sub-arc parental compositions. However, new olivine (n = 6) and clinopyroxene (n = 12)  18O values obtained by Laser Fluorination (LF) analysis display a narrow range, with averages at 6.0‰ ± 0.2 (2σ S.D.) and 6.7‰ ± 0.3 (2σ S.D.), consistent with a common history for the investigated minerals. Additionally, the first in-situ Secondary Ionisation Mass Spectrometry (SIMS) oxygen isotope data for mafic mineral phases from La Poruña and Palpana is presented in order to ensure a robust parental melt δ18O composition, since unlike in LF analysis fractures or mineral/melt inclusions can be avoided. Crystal cores show a density peak at 5.6‰ for both La Poruña and Palpana, whereas rims have either relatively high (6.0‰) or low (5.4‰) peaks. Intra-crystal differences (up to ± 1.1‰) in  18O values between crystal cores and rims accurately record crustal assimilation that postdate cores formation at 13 to 19 km depth. Based on the more robust dataset of LF results, a  18O value for sub-arc, parental melts of ca. 7.0‰ ± 0.2 (2σ S.D.) can be estimated. SIMS core data, on the other hand, indicate a lower, but less well-constrained parental melt value of 6.3‰ ± 0.4 (2σ S.D.).Parental melts feeding the APMB and associated volcanic centres are proposed to form in the felsic continental crust following assimilation of up to 28 vol% high  18O-high 87Sr/86Sr basement rocks by mantle-derived magmas.

No description available.

Cathodoluminescence photomicrographs of diamonds from the George Creek Kl (section 28) kimberlite dyke in Colorado reveal complex intergrowth relationships between CO?-free and CO?-bearing diamond growth generations. The distribution of the CO?-bearing diamond in some specimens suggests that this generation is younger than the CO?-free diamond growth generation, although the age relationships are mostly ambiguous. CO?-bearing diamond appears to have crystallized from fluids which invaded fractures and etched embayments in the CO?-free diamond growth generation, which shows evidence of plastic deformation. The CO?-free diamond growth generation commonly exhibits features caused by extreme plastic deformation during mantle residence time. Abundant yellow-green plastic slip planes transect zones of customary blue cathodoluminescence in many diamonds, and raised lamination lines have been recognized on resorption surfaces. The complexity and intensity of surface etch features in most George Creek diamonds, including the CO?-bearing growth generation, suggests that the diamonds were subjected to multiple episodes of etching and resorption. Extensive development of hexagonal and trigonal etch pits resulted from the action of oxidizing CO?-H?O fluids, and some late-stage etching is believed to have occurred in the hypabyssal dyke system prior to kimberlite eruption.

The most extensive evidence for the evolution of the primitive crust is preserved at the base of the continental roots (140 to 200 km depth). However, this is completely inaccessible and hard to evaluate, except through the study of mantle xenoliths brought to the surface by kimberlite magmas. Most mantle xenoliths undergo kimberlitic/carbonaceous metasomatism prior to their entrapment, altering their primary composition. Despite a very complex history and a very low abundance (180 nodules), sampled in four main localities from the Siberian (Obnazhennaya, Udachnaya) and South African (Jagersfontein, Roberts Victor) cratons. Major and trace element compositions have been analysed in representative bimineralic (garnet (gt)-omphacite (cpx)), coesite-, kyaniteand corundum-bearing eclogites. Metasomatized eclogites (Type I) have a higher alkali- (Na2O in garnet and K2O in omphacite) and LREE, LILE-content. Non-metasomatized (Type II) eclogites include two main compositional groups, derived from different protoliths. The first group (Type IIA) is typically more magnesian, enriched in LREE and has δ18O values from 3.73 to 7.50 ‰, with positive Sr anomalies and corresponding to a low pressure-temperature layer. The whole-rock composition is consistent with a basaltic protolith. The second group (Type IIB) is more sodic and aluminous, depleted in LREE and has δ18O values of 2.35 to 3.59 ‰, corresponding to equilibrium at high pressure and temperature. The whole-rock trace element composition is consistent with a pyroxenitic protolith. Eclogites that contain coesite, kyanite and corundum (coe-ky-cor) are typically characterized by jadeite-rich clinopyroxenes with positive Eu and Sr anomalies and grossular-rich garnets with corresponding positive Eu and negative Sr anomalies. Additionally, corundum-bearing samples are overall LREE-depleted. Pressure-temperature estimates indicate coe-ky-cor-bearing eclogites equilibrated in the lowermost part of the cratonic keel, and the reconstructed whole-rock trace element composition corresponds to a very depleted gabbroic protolith. Corundum-bearing eclogites often show topotactic exsolution textures hosted in omphacite, consisting of a more calcic garnet ± kyanite/zoisite. Based on Si, Al and Mg diffusion profiles it can be inferred that exsolution was controlled by chemical exchange in an isochemical system. Similarly, an interpretation can be made for the variable HREE fractionation of intermediate composition between the exsolutions and the host omphacite, as evidence for intermineral partitioning. The formation of zoisite (1.7 % H2O) from a precursor clinopyroxene requires a significant amount of hydrogen (as H+ , OH- or H2O) incorporated in omphacite at mantle conditions. Calculated water content of omphacite, based on Fourier transform infrared spectrometry (FTIR) analyses, varies from ~930—1410 to ~1100—1680 ppm by weight H2O, according to different absorption coefficients. Primary garnets are typically anhydrous (<7 ppm H2O), whereas coarse exsolutions contain between 165—1950 ppm H2O. Reconstructed estimates for whole-rock water content (~310—890 ppm H2O) for the Obnazhennaya eclogites are significantly higher than those of the surrounding peridotites. The variability of δ18O in garnet among 41 xenoliths, shows a bi-modal distribution with median values at 3.57 ‰ and 5.68 ‰ and strong correlation (r = 0.96) between garnet and omphacite. The δ18O values and the reconstructed whole-rock trace element compositions indicate an oceanic crustal protolith for mantle eclogite xenoliths. This is consistent with the subduction of a hydrothermally altered, basaltic to websteritic sequence of an incompatible-element-depleted oceanic crust. Furthermore, although eclogites represent a small fraction of the upper mantle, they may be an important water reservoir at the base of the cratonic root. High water content in the lithospheric mantle would have major consequences for the longevity of the cratonic keel, forphysical and chemical properties in minerals, partial melting, mantle rheology and electrical conductivity and the global water cycle.