层状岩体的成因及成矿作用

层状岩体的典型特征是其具有层状构造和韵律结构。层状镁铁质岩体不仅是研究岩浆分异演化过程和地壳混染作用的最佳天然实验室,其中还广泛赋存重要的稀有金属和贱金属矿床。文中主要总结了不同类型层状岩体形成的构造背景、基本特征、母岩浆特征和韵律层理成因,以及岩浆混合和地壳混染在岩浆演化过程中起到的重要作用。文中还介绍了层状岩体中典型矿床的成矿作用与岩浆演化过程的成因联系。     

The cumulus and post-cumulus evolution of chrome-spinels in ultrabasic layered intrusions: evidence from the Cuillin Igneous Complex,Isle of Skye,Scotland

Chrome-spinels from the layered Peridotilte Series of the unmetamorphosed, anorogenic 60 Ma Cuillin Igneous Complex, Isle of Skye, display a wide variety of compositions. Cumulus (within seams) chrome-spinels from the lowest exposed portion of the Peridotite Series exhibit features indicative of textural equilibrium, are rich in Al and Mg, and have low values of the ratio Cr/(Cr+Al). Cumulus chrome-spinels from higher up in the series are different from these: particularly, textural disequilibrium is evident, intercumulus plagioclase and olivine are present, and the chrome-spinels are rich in Cr, Fe and Ti, with high values of the ratio Cr/(Cr+Al). Intercumulus (dispersed) chrome-spinels tend towards anhedral forms and define enrichment trends towards Fe (both Fe²⁺ and Fe³⁺) with decreasing Mg, Cr and Al, and towards Al, with decreasing Fe²⁺ and Cr (and increasing Mg). Individual crystals are completely homogeneous and are devoid of reaction rims. The observed textural characteristics and compositional data of the chrome-spinels documented here suggest that the semi-quantitative peritectic reaction: aluminous chrome-spinel + meltplagioclase + olivine + chromian chrome-spinel, is responsible for the observed parageneses, and that both the environment of crystallization (eumulus or intercumulus) and the role of plagioclase ±olivine crystallization are critical parameters for this geochemical trend in spinels within upper crustal magmatic systems. The effects of pyroxene crystallization on the development of this geochemical trend are also considered. This investigation highlights the need to consider the role of post-cumulus mineral-melt reactions and their influences upon the final compositions of major oxide and silicate phases within layered intrusions.     

The high pressure experimental crystallization of staurolite in hydrous mafic compositions

The experimental crystallization of staurolite in two mafic compositions at high pressures (24–26 kb) and temperatures (740 °–760 ° C) is reported. This, together with a recent report of staurolite from metamorphosed basic igneous rocks from Fiordland, New Zealand, leads to the conclusion that Staurolite may be an important subordinate phase in the subducted lithosphere. The synthesized staurolites are more magnesian than those occuring in natural metapelitic assemblages and are characterized by relatively high Si, Ti, Mg, Ca, and low Al, Fe.     

Evolution of major mineral compositions and trace element abundances during fractional crystallization of a model lunar composition

The evolution of major mineral compositions and trace element abundances during fractional crystallization of a model lunar magma ocean have been calculated. A lunar bulk composition consistent with petrological constraints has been selected. Major mineral compositions have been calculated using published studies of olivine-melt, plagioclase-melt, and pyroxene-olivine equilibria. Trace element abundances have been calculated using experimentally-determined partition coefficients where possible. In the absence of experimental determinations, published partition coefficients obtained by analyzing phase separates from porphyritic volcanic rocks have been used. Trace elements studied are La, Sm, Eu, Lu, Rb, Sr( Eu²⁺), Ni, Co, and Cr.The first mineral to crystallize is olivine, which varies in composition from Fo₉₈ at the liquidus to Fo₉₅ at 50% solidification. Orthopyroxene crystallizes from 50 to 60% solidification with a restricted composition range of En₉₅-En₉₃. Plagioclase and Ca-rich clinopyroxene (X_Wo arbitrarily set equal to 0.5) co-crystallize during the final 40% solidification. Plagioclase changes in composition from An₉₇ to approximately An₉₃, while clinopyroxene evolves from En₄₆ to approximately En₄₀. The concomitant evolution of major element abundances in the melt is also discussed.The concentration of Ni in the melt decreases rapidly because solid-melt partition coefficients are significantly greater than unity at all stages of crystallization. The concentration of Cr in the melt increases slowly during olivine crystallization, then drops precipitously during the crystallization of orthopyroxene and clinopyroxene. The concentration of Co in the melt decreases slowly during olivine and orthopyroxene crystallization, after which it returns slowly to its initial concentration. Rubidium and Sr are not fractionated relative to one another until the onset of plagioclase crystallization. Ratios of Rb/Sr, normalized to their initial concentrations in the magma, do not rise above 10 until 95% of the magma has solidified. The ratios of Eu/Sm and La/Lu, normalized to their initial concentrations in the magma, remain essentially unfractionated until the onset of crystallization of clinopyroxene plus plagioclase, at which point the normalized La/Lu ratio increases to approximately 1.3 at 100% solidification and the normalized Eu/Sm ratio decreases to approximately 0.2 at 100% solidification.The model calculations are used to place approximate constraints on the bulk composition of the primitive Moon. Consideration of the effect on plagioclase composition of the activities of NaO_0.5 and SiO₂ in the melt suggests that the primitive Moon contained less than 0.4 wt % NaO_0.5 and approximately 42–43 wt % SiO₂. Concentrations of the REE in model lunar anorthosites are consistent with the returned samples. Concentrations of the REE in several model ‘highland basalts’ (considered to be representative of the average lunar terrae) are too low when compared with returned samples. Several possible explanations of this discrepancy are considered. The possible role of spinel in a twostage geochemical evolution of mare basalt liquids is discussed.     

攀西裂谷地区层状镁铁岩的PGE矿化作用

攀西裂谷位于四川西部,裂谷经历了元古宙和海西期二次地幔柱活动,形成多处穹窿构造和层状镁铁质岩体的侵入。后一期的层状岩体赋存著名的超大型钒钛磁铁矿床。中国和南非合作研究认为,层状岩体PGE矿化应进一步研究。以新街岩体为代表,经钻探工程建立了岩体剖面;岩石学、矿物学和地球化学研究证实,岩体有三个岩浆旋回和许多韵律层,层厚仅2~3cm。自上而下岩相为辉长岩、橄辉岩、辉石岩和橄榄岩,造岩矿物为贵橄榄石、普通辉石、钛普通辉石和中长石。岩体下部旋回,硫化物较富集,多在高镁质岩相。硅酸盐、氧化物和硫化物三系列矿物共生而不混熔。硫化物呈浸染状,主要有三层,产在橄榄岩、辉石岩和下辉长岩内。铂族矿物有砷铂矿、自然铂、硫锇矿、铋碲钯矿、碲铋矿、碲银矿、自然银等。PGE富集可能有三个阶段:岩浆早期,岩浆中"S"不饱和,PGE易进入硅酸盐;岩浆晚期"S"逸度增高,硫化物富集,为PGE富集阶段;热液阶段PGE再分配富集。PGE和Ni、Cu、S为正相关关系,和Fe、Ti相辅相成,无明显关系。岩石中PGE背景值为(0.166~0.411)×10-6,PGE矿化体的品位变化较大,为(0.94~0.976)×10-6。有的钻孔样品Pt+Pd含量大于1×10-6,可做进一步找矿的依据。     

Crustal assimilation in basalt and jotunite: Constraints from layered intrusions

To constrain the amount and rate of crustal contamination that is possible in basaltic and jotunitic magma, and to gain an insight into the physical and thermal processes of assimilation in crustal magma chambers, we have modelled published Sr and Nd isotopic data from three layered intrusions. Well-exposed sequences of cumulates with no evidence of magma recharge provide direct records of concurrent assimilation and fractional crystallization (AFC). The key to the modelling is that F, the mass fraction of magma remaining in the chamber, can be estimated from the thicknesses of the studied cumulate sequences. This allows AFC model curves to be fitted to the isotopic data by varying r, the ratio of the rate of mass assimilated to the rate of mass crystallized. The results of modelling show that r is nearly constant in 800 to 2000 m thick sequences of cumulates displaying up-section decreases in anorthite content of plagioclase, increases in whole-rock Sr₀ (initial ⁸⁷Sr/⁸⁶Sr) and decreases in whole-rock εNd₀ (initial εNd). The r-values of the layered sequences range from 0.12 in the Fongen–Hyllingen Intrusion, over 0.20 in the Bjerkreim–Sokndal Intrusion, to 0.27 in the Hasvik Intrusion. The total amount of assimilation, the bulk crust/magma ratio, reaches values of 0.08, 0.19 and 0.28 at the level of the most contaminated samples after 60% to 80% crystallisation, whereas the instantaneous crust/magma ratio of the most contaminated magmas were respectively 0.14, 0.46, and 0.70, for the three intrusions.Innumerable country rock xenoliths occur in the three layered intrusions and played a crucial role in the assimilation process. The xenoliths spalled off the roofs of the magma chambers during magma emplacement and their initial temperature and composition relate to r in the intrusions. In the Hasvik Intrusion (r = 0.27), the initial temperature of the country rocks was 450 °C and the xenoliths were fusible metasediments and therefore produced a high fraction of partial melt that could be assimilated. In the Bjerkreim–Sokndal Intrusion (r = 0.20), the country rocks were initially at temperatures of 640–880 °C but included both refractory massif-type anorthosite and fusible gneisses. In the Fongen–Hyllingen Intrusion (r = 0.12), the country rocks were cooler (300 °C) and the xenoliths include refractory metabasalt (dominant) and fusible metapelite. We argue that the refractory metabasalt and anorthosite xenoliths acted mainly as heat sinks, resulting in reduced r-values in Fongen–Hyllingen and Bjerkreim–Sokndal Intrusions.Heating of refractory and fusible xenoliths, and melting of fusible xenoliths absorbed sensible and latent heat of the magma. Energy-balanced modelling shows that up to 75% of the heat available was absorbed by xenoliths within the magma chambers, promoting higher rates of cooling and crystallisation than would have resulted from loss of heat to the envelope of country rocks alone. The high r-values reflect the amount of heat absorbed by heating and melting country rock within the magma chambers themselves, and their constancy reflects the ready availability of fusible xenoliths.     

Crustal assimilation in basalt and jotunite: Constraints from layered intrusions

Massive continental volcanism and/or bolide impacts are considered by many authors to have caused three major mass extinction events during the last 300 million years: the end-Permian, end-Cretaceous and end-Triassic extinctions. However, re-evaluation of the frequency of bolide impacts and plume-related flood basalt provinces indicates that both types of event occur much more frequently than mass extinctions, and so, in isolation, may not be responsible for the largest extinctions. Furthermore, the kill mechanisms associated with either flood basalts or impacts do not appear to be sufficiently powerful to cause worldwide collapse of ecosystems leading to the largest mass extinctions. Contemporaneous flood basalts and bolide impact may be prerequisites for the largest mass extinctions. We present a statistical analysis of the probability of coincidence between volcanism and impact, and show that three random coincidences of these events in the last 300 m.y. are likely. No causal relationship between impact and volcanism is necessary. The lesser mass extinctions, on the other hand, may not require juxtaposition of two such catastrophic events; such coincidences occurring on more than three occasions during the last 300 m.y. become increasingly unlikely.     

Early Proterozoic layered intrusions in the northeastern part of the Fennoscandian Shield

Summary Early Proterozoic layered intrusions, about 2440 Ma in age, are widespread over a large area of the northeastern Fennoscandian Shield in Finland, Sweden and the Soviet Union. Only one intrusion, the Kukkola intrusion, is encountered in Sweden whereas in Finland, their number exceeds twenty. These are concentrated principally in two areas, the dicontinuous Tornio-Närdnkävaara intrusion belt which crosses northern Finland and the Koitelainen intrusion with its satellites located in central Finnish Lapland. The intrusions in the Soviet Union are concentrated in three areas: (i) on the Kola Peninsula, (ii) in the Paanajärvi area close to the Finnish border and (iii) northeast of Lake Onega.Examples of all the ore types characteristic of layered intrusions have been found in these intrusions. Chromitite layers are encountered in the Kukkola/Tornio, Kemi, Penikat, Koitelainen and Burakovsky intrusions, but only one, the Kemi chromitite, has so far been mined. The Portimo, Koillismaa, Monchegorsk and Fedorova intrusions are characterized by PGE-bearing Cu-Ni-deposits in their marginal series. Mineralized zones enriched in PGE are also encountered in the layered series. Those in the Penikat intrusion and in the Portimo intrusions are the most remarkable and the best known to date. Vanadium-bearing Fe-Ti-oxide layers are encountered in several intrusions, but only one, the Mustavaara deposit, is presently being exploited.Two types of parental magma have tentatively been proposed for these intrusions. The first type is represented by a magma which was relatively rich in magnesium and chromium and was as a whole boninitic in composition, whereas the plagioclase-rich intrusions and megacyclic units are interpreted as having crystallized from a magma which was greatly depleted in these elements, especially Cr, and had melted crustal material incorporated in it.The emplacement of the early Proterozoic layered intrusions in Fennoscandia was part of the world-wide igneous activity indicated by other layered intrusions and mafic dyke swarms of similar age in other ancient cratons, i.e. the Jimberlana intrusion in Australia, the Great Dyke in Zimbabwe, the Scouric picrite suite in Scotland, the Hearst-Matachewan dyke swarm, Copper Cliff Formation and East Bull Lake intrusion in Ontario, Canada, and the Vestfold Hills and Napier Complex dyke swarms in Antarctica. This almost contemporaneous occurrence in different parts of the world would suggest a more intimate relationship between the Fennoscandian Shield, northwest Scotland, Canadian Shield, Yilgarn Block, Zimbabwe Craton and East Antarctic Shield at the beginning of the Proterozoic than at present.

---

Früh-Proterozoische geschichtete Intrusionen im nordöstlichen Teil des Fennoskandischen Schildes

Zusammenfassung Im Nordost-Teil des Fennoskandischen Schildes in Finnland, Schweden und der Sowjetunion kommen fast vierzig frühproterozoische geschichtete Intrusionen, die ungefähr 2440 Mio J. alt sind, vor. Nur eine davon, die Kukkola Intrusion, liegt in Schweden, während in Finnland mehr als zwanzig Intrusionen vorkommen. Diese sind hauptsächlich in zwei Gebieten konzentriert, nämlich in dem nicht-zusammenhängenden Tornio-Näränkävaara Gürtel, der das nördliche Finnland durchzieht, und die Koitelainen-Intrusion mit ihren Satelliten im zentralen Finnischen Lapland. Die Intrusionen in der Sowjetunion sind in drei Gebieten konzentriert: (i) auf der Kola Halbinsel (ii) im Paanajärvi Gebiet nahe der Finnischen Grenze und (iii) östlich vom Onega-See.Beispiele aller für geschichtete Intrusionen charakteristischen Erztypen kommen vor. Chromititlagen sind in den Intrusionen von Kukkola/Tornio, Kemi, Penikat, Koitelainen und Burakovsky zu finden, aber nur eine davon, der Kemi Chromitit, ist bisher in Abbau genommen worden.Die Portimo-, Koillismaa-, Monchegorsk- und Fedorova-Intrusionen werden durch PGE-führende Kupfer-Nickel-Lagerstätten in ihren randlichen Bereichen charakterisiert. Mineralisierte Zonen die an PGE angereichert sind kommen auch in den geschich teten Serien vor. Die bemerkenswertesten sind die PGE-Vererzungen der Penikat- und der Portimo-Intrusionen. Vanadium-führende Fe-Ti-Oxidlagen kommen in verschiedenen Intrusionen vor, aber nur eine davon, die Mustavaara-Lagerstätte, ist bisher abgebaut worden.Diese Intrusionen werden auf zwei verschiedene Magmentypen zurückgeführt. Ersteres ist ein Magma das relativ reich an Magnesium und Chrom war und eine boninitische Zusammensetzung hatte, während die Plagioklas-reichen Intrusionen, und die megazyklischen Einheiten auf ein Magma das an diesen Elementen (besonders Cr) verarmt war, und das Krustenmaterial aufgeschmolzen hat, zurückgehen.Die Platznahme der frühproterozoischen geschichteten Intrusionen in Fennoskandien stellt einen Teil weltweiter magmatischer Aktivität dar, die durch andere geschichtete Intrusionen und mafische Gänge von fast identischem Alter in anderen alten Kratonen repräsentiert wird. Hier ist die Jimberlana-Intrusion in Australien, der Great Dyke in Zimbabwe, die Pikrit-Suite von Scourie in Schottland, die Gänge von Hearst-Matachewan, die Copper Cliff Formation und die East Bull Lake Intrusion in Ontario, Kanada ebenso wie die Gangsysteme der Vesthold Hills und des Napier Komplexes in Antarctica zu nennen. Diese fast gleichaltrigen Vorkommen in verschiedenen Teilen der Welt weisen auf eine engere Beziehung zwischen dem Fennoskandischen Schild, Nordwest-Schottland, dem Kanadischen Schild, dem Yilgarn Block, dem Zimbabwe-Craton und dem Ostantarktischen Schild zum Beginn des Proterozoikums hin.

---

---

With 7 Figures     

Phase equilibrium studies of the chilled margins of some layered intrusions

The chilled margins of the Michikamau, Muskox, Skaergaard, and Bushveld intrusions were experimentally remelted over a range of f _{O 2}. The observed primary phases, at certain values of f _{O 2} (generally 10^–12 to 10^–8 atm) correspond with the reported mineralogy of the earliest cumulates. Previous investigations appear to have been conducted at higher values of f _{O 2} (>10^–6 atm) at which the experimental mineralogy did not relate to the cumulates.October 1973 to October 1974 at Division of Tribophysics, C.S.I.R.O., P.O. Box 4331, Melbourne, Victoria 3001, Australia.     

Archaean layered intrusions of the Eastern Goldfields region,Western Australia

Small Archaean, pre-metamorphic layered sill-like intrusions showing Stillwatertype differentiation are widespread in the Eastern Goldfields region of Western Australia. The sills have undergone low-grade metamorphism, but the retention of some original minerals and excellent preservation of primary textures have enabled most of their original features to be reconstructed. Despite their small size in comparison with well known layered intrusions, simple differentiation sequences are developed ranging upwards from dunite and harzburgite through orthopyroxenite, norite, norite-gabbro and gabbro to granophyre and anorthosite. The parent magma for the sills appears to have been rich in MgO (15–18%) and is probably represented by the spatially associated high-Mg basalts found in the region. The complexes are interpreted as being contemporaneous with volcanism and may be high-level emplacements.     

攀西三个镁铁-超镁铁质岩体的地球化学及其对源区的约束

在峨眉山大火成岩省的攀西地区的南北向断裂带内分布着一系列层状镁铁-超镁铁质岩体,它们在时空上与峨眉山玄武岩有着密切的联系,代表了岩浆管道系统.本文选择了3个代表性岩体,含钒钛磁铁矿的攀枝花岩体,含铜镍硫化物矿床的力马河岩体以及同时舍钒钛磁铁矿和铜镍硫化物铂族元素矿床的新街岩体进行了主要元素、微量元素和Sr-Nd同位素地球化学研究.总体上新街岩体在地球化学特征上均介于攀枝花岩体和力马河岩体之同.在主要元素特征上,攀枝花岩体相对富Fe、Ti、Ca、Al,而力马河岩体则相对富Mg.3个岩体的稀土和微量元素标准化由线均与丽江苦橄岩相似,均显示出LREE和LILE的相对富集,但是力马河岩体有明显的负Nb、Ta和Ti异常,反映了源区有地壳物质或岩石圈地幔的混染.在log(Th/Yb)PM vs. log(Ta/Th)PM图解上,3个岩体昱示出线性关系,其中攀枝花岩体具有高的(Ta/Th)PM和低的(Th/Yb)PM值,而力马河岩体则显示出相反的特征,并且该图解也暗示了攀枝花岩体有岩石圈地幔的混染,力马河岩体有上地壳物质的混染,而新街岩体则有下地壳物质的混染.在Sr-Nd同位素特征上,3个岩体也有明显的区别,其中新街岩体具有较均一的(87Sr/86Sr)t(0.70610～0.70636)和εNd(t)值(1.01 to 1.75),力马河岩体则具有高的(87Sr/86 Sr).(0.70631～0.70930)和低的εNd(t)值(-0.85 to -4.13),而攀枝花岩体则与丽江苦橄岩的Sr-Nd同位素组成(0.70434～0.70510,1.1～3.2)相似.然而,3个岩体的Sr-Nd同位素成分(除力马河岩体部分样品外)均落在峨眉山玄武岩和洋岛玄武岩(OIB)内,因此这3个岩体可能均来自于地幔柱,但地幔柱熔体在上升过程中受到了不同物质的混染,其中,攀枝花岩体的源区为地幔柱和岩石?圈地幔,新街岩体为地幔柱和下地壳,而力马河岩体则为地幔柱和上地壳,其源区的特征可能反映了地幔柱熔体的演化过程.     

The distribution of phosphorus in the early and middle stages of fractionation of some basic layered intrusions

The distribution of phosphorus in those rocks without primary apatite from the Skaergaard, Bushveld and Rhum layered intrusions has been determined. The concentration of phosphorus in the cumulus phases is very low but in some of the rocks nearly all the phosphorus can be accounted for by the amount in these minerals. In all three intrusions the order of entry of phosphorus into the cumulus minerals is olivine >; pyroxene > plagioclase feldspar. Some of Koritnig's conclusions concerning the distribution of phosphorus are criticised. Phosphorus can act as an indicator of the relative amounts of pore material in basic cumulate rocks without primary apatite, provided correction is made for the phosphorus in the cumulus phases and that concentrations are expressed in weight per unit volume.     

The calculation of mineral compositions and modes of olivine-two pyroxene-spinel assemblages

Duhem's theorem states that the equilibrium position of a system of known bulk composition and size is completely determined by two independent variables. Given these data all other equilibrium parameters can be calculated. The method of calculation is set down in detail for olivine-two pyroxene-spinel assemblages which may serve as models for portions of the upper mantle. To make the calculations, equations for the Gibbs energy of the constituent minerals, olivine, spinel, orthopyroxene and clinopyroxene are formulated. The olivine is treated as an ideal, binary, cationic solid solution. The spinel is treated as a ternary, reciprocal solution between the four end members MgAl₂O₄, MgCr₂O₄, FeAl₂O₄ and FeCr₂O₄. The pyroxenes are treated as seven-component, simplified reciprocal systems. The eight oxides SiO₂, TiO₂, Al₂O₃, Cr₂O₃, FeO, MgO, CaO and Na₂O are thought sufficient for expressing the composition of the model mantle.The paucity and inconsistency of the required thermodynamic data preclude accurate calculations. To illustrate the calculation and to arrive at some highly tentative results, the required free energy data are estimated by combining analytical data from Iherzolite xenoliths with interpretations of experimental data.It seems likely that the Al content of pyroxenes cannot be used as a geobarometer without considering the bulk composition of the assemblage or, equivalently, the composition of the associated spinel phase. Temperature estimates based on the Ca to Ca+Mg ratio in Ca-rich pyroxenes may be approximately correct in the restricted composition range of mantle derived xenoliths.     

Evidence for the trapped liquid shift effect in the Mount Ayliff Intrusion,South Africa

Postcumulus trapped liquid shift in layered complexes produces cumulate minerals with more fractionated compositions than the original primary phases. This effect is shown by olivine compositions from the base of the Mount Ayliff Intrusion, where varying proportions of olivine to interstitial liquid produce a suite of rocks which define a tight linear trend on a binary whole-rock plot of MgO versus FeO. Extrapolation of this trend constrains the composition of the primary cumulus olivine to the range Fo_84–86, whereas olivine actually present have compositions Fo_77–83. The magnitude of the discrepancy between the theoretical and observed olivine compositions correlates directly with the weight fraction of interstitial liquid. These observations are quantitatively predicted by the trapped liquid shift model. They also argue against significant migration of residual liquid. Trapped liquid shift is documented over a vertical interval of 60 m. It occurred in rocks lying only 1 m above the basal contact of the intrusion and hence must be a comparatively rapid process.     

Modeling incompatible trace-element abundances in plagioclase in the Skaergaard intrusion using the trapped liquid shift effect

Incompatible trace-element abundances in minerals and whole rocks from layered intrusions have been used to model the fractionation processes and evolving liquid compositions. Many such models assume that the analyzed concentration in a mineral represents that of the mineral when it first crystallized. However, overgrowth from residual liquid and subsequent diffusive equilibration can result in significant changes to the bulk mineral compositions (the more incompatible the element the more dramatic the subsequent changes). The proportion of that residual liquid relative to the cumulus minerals is the most important parameter in determining the magnitude of this effect (trapped liquid shift effect). Calculations involving Ba and La contents in plagioclase quantitatively demonstrate this effect. For Ba and La (partition coefficients of 0.4 and 0.04), 50% trapped liquid in a sample can result in two and sevenfold increases, respectively, in concentration between original and final bulk mineral compositions. Different cumulus assemblages also have a major effect on final compositions. We use examples of the concentrations of Ba and La in plagioclase from the Skaergaard intrusion from previous publications to illustrate the importance of this effect. Specifically, the La content of bulk plagioclase steadily decreases upward from the Lower Zone to Upper Zone c, and Ba in plagioclase shows no increase from the Lower Zone to the top of the Middle Zone. Such results are not explicable by fractionation processes, but can be modeled by the trapped liquid shift effect, assuming the well-established evidence for upward decrease in trapped liquid proportion through these zones.     

The effect of Mn on mineral stability in metapelites

Calculations based on a KMnFMASH petrogenetic grid derived using an internally consistent thermodynamic dataset indicate that the principal effect of the presence of Mn in average subaluminous pelite compositions is to stabilize garnet to higher and lower pressures and temperatures over a wide range of bulk compositions. Garnet-bearing fields expand to lower temperatures and pressures with the addition of Mn, and garnet appears as an extra phase at low pressures. The addition of Mn also increases the number and extent of four AMnFM phase assemblages and stabilizes five AMnFM phases along univariant reactions. The KMnFMASH system predictions for typical subaluminous pelite bulk compositions match the sequence of isograds and assemblages observed in the Barrovian zones. The sequence of assemblages observed in the Stonehaven section can also be predicted if there is variation in bulk composition within the stratigraphic section. Mn appears to be less important in producing the sequence of isograds and garnet-absent assemblages in the low-pressure Buchan zones. The addition of Mn to the calculations does not change the sequence of isograds that are predicted to be stable in a regional metamorphic terrane, but the P – T  position of these isograds does change. In particular, the predicted temperature of the garnet-in isograd is lowered by as much as 100 °C by the addition of Mn to KFMASH. Mn also increases the range of metapelite bulk compositions that develop the assemblages traditionally identified as metapelite isograds.     

山东中侏罗世-早白垩世侵入岩的锆石Hf同位素组成及其意义

对山东中侏罗世-早白垩世侵入岩中锆石的原位Hf同位素分析显示,形成于晚太古代（上交点年龄～2.5 Ga）的继承锆石具有正的εHf（t）值（＋8～＋1）,Hf同位素模式年龄集中在2.6～2.8 Ga,与辽宁古生代金伯利岩中基性下地壳捕虏体中锆石Hf组成和Hf模式年龄十分一致,Hf模式年龄也与研究区变质岩和花岗岩的全岩Nd模式年龄相同,因此,这些继承锆石来自于晚太古代由岩浆底侵形成的基性下地壳.新生锆石出现在继承锆石周围或者以独立颗粒出现,其U-Pb年龄为177 Ma和132～126 Ma,εHf（t）值均为负值（-23～-1）.山东中生代侵入岩的形成与富集岩石圈地幔,亏损地幔和地壳三个端员之间的相互作用有关.其中根据来源于晚太古代下地壳的侏罗纪铜石二长花岗岩限定的研究区下地壳εHf（t）平均值为-20,根据来源于富集岩石圈地幔的早白垩纪沂南辉长岩限定的富集地幔端员的εHf（t）为-16.部分样品锆石εHf（t）变化非常大（-20～-1）,示踪了岩浆作用过程中亏损地幔物质的参与程度的逐渐增强.这种变化是华北晚中生代岩石圈大规模减薄作用的结果.     

The effect of crystallization time on plagioclase grain shape in dolerites

The average aspect ratio of plagioclase, measured in thin section, varies systematically through dolerite sills, with a symmetrical “M”-shaped profile observed in sills thinner than ~200 m. Thicker sills show the same marginal reversal at the base, but average aspect ratios appear to continue increasing towards the top, creating an “S”-shaped profile. A distorted “M”-shaped profile is visible in the stratigraphic variation of plagioclase average aspect ratios in the prehistoric Makaopuhi lava lake, with the centre of the “M” in the olivine-enriched horizon close to the base of the lake. Higher overall values of average aspect ratio are observed in thin sills compared to thicker sills, demonstrating that the plagioclase in more slowly crystallized bodies is more equant than that in more rapidly crystallized bodies. There is a strong correlation between the plagioclase average aspect ratio in the central parts of the sills and the crystallization time calculated using a simple one-dimensional thermal model assuming conductive cooling. The cause of the marginal reversals in average aspect ratio is not well understood, but may result from early grain impingement in the developing crystal mushy layer.