Search

The chemistry of surviving pieces of Eoarchaean mantle together with related crust helps us determine early crust-forming mechanisms. Two lenses of high-Mg, low-Al dunite within a ca. 3720 Ma part of the Isua supracrustal belt in Greenland are interpreted as relicts of Eoarchaean mantle with minimal crustal disturbance. The lenses are within altered, higher Al, Ca ultramafic schists and are intercalated with amphibolitized pillow basalts and gabbros with island arc chemical signatures, all intruded by 3715–3710 Ma tonalites. One variety of dunite is dominated by forsterite (Fo90–92) olivine with accessory chromite and rare clinopyroxene, which does not show high field strength element (HFSE) anomalies. Another variety contains olivine (Fo96–98), some intergrown with Ti-humite group minerals with strong positive HFSE anomalies that are complementary to the negative HFSE anomalies of the adjacent amphibolites. We propose that these dunites are tectonic slivers of ca. 3720 Ma subarc mantle that preserve evidence for varying interaction with mafic magmas in a ≥850 °C, 1.7–2.0 GPa subcrustal environment. These are by far the oldest direct geochemical link between coeval mantle and crustal rocks, and are new evidence for subduction zone–like environments on the early Earth.

New SHRIMP U–Pb zircon geochronology and fieldwork integrated with reappraisal of earlier mapping demonstrates that the so-called ‘southern region’ of the mainland Lewisian Gneiss Complex comprises a package of distinct tectono-stratigraphic units. From south to north these are the Rona (3135–2889 Ma), Ialltaig (c. 2000 Ma) and Gairloch (ca. 2200 Ma) terranes. These terranes were metamorphosed and deformed separately until ca. 1670 Ma by which time they had been juxtaposed and were integral with terranes to the north. The northern boundary of the Palaeoproterozoic Gairloch terrane is a shear zone, north of which is the Archaean Gruinard terrane with 2860–2800 Ma protoliths and ca. 2730 Ma granulite facies metamorphism. In contrast, south of the Gairloch terrane, the Archaean gneisses of the Rona terrane have older protolith ages, underwent an anatectic event at ca. 2950 Ma and show no evidence of 2730 Ma granulite facies metamorphism. In current structural interpretations the Gruinard terrane forms a structural klippe over the intervening Gairloch terrane. However, the Rona and Gruinard terranes cannot be equivalent on age grounds, and are interpreted as unrelated different entities. Contained within the southern margin of the Gairloch terrane is the Ialltaig terrane, shown here to comprise an exotic slice of granulite facies Palaeoproterozoic crust, rather than Archaean basement as previously thought. The ca. 1877 Ma granulite facies metamorphism of the Ialltaig terrane is the youngest event that is unique to a single terrane in the mainland Complex, making it an upper estimate for the timing of amalgamation with surrounding tectonic units. U–Pb titanite ages of 1670 ± 12 Ma and ca. 1660 Ma for low-strain zones at Diabaig are interpreted to be cooling through the titanite closure temperature after the amphibolite facies reworking of these southern terranes and the southern margin of the Gruinard Terrane. These new data have implications for the tectonic setting of the mainland in relation to the Outer Hebrides and in the wider evolution of the basement in the North Atlantic.

In situ LA-ICP-MS monazite geochronology from a garnet-bearing diatexite within the Moine Supergroup (Glenfinnan Group) NW Scotland records three temporally distinct metamorphic events within a single garnet porphyroblast. The initial growth of garnet occurred in the interval c. 825–780 Ma, as recorded by monazite inclusions located in the garnet core. Modelled P–T conditions based on the preserved garnet core composition indicate an initially comparatively high geothermal gradient regime and peak conditions of ∼650 °C and 7 kbar. Monazite within a compositionally distinct second shell of garnet has an age of 724 ± 6 Ma. This is indistinguishable from a SIMS age of 725 ± 4 Ma obtained from metamorphic zircon in the sample, which is interpreted to record the timing of migmatization. This second stage of garnet growth occurred on a P–T path from ∼6 kbar and 650 °C rising to ∼9 kbar and 700 °C, with the peak conditions associated with partial melting. A third garnet zone which forms the rim contains monazite with an age of 464 ± 3 Ma. Monazite in the surrounding matrix has an age of 462 ± 2 Ma. This corresponds well with a U–Pb SIMS zircon age of 463 ± 4 Ma obtained from a deformed pegmatite that was emplaced during widespread folding and reworking of the migmatite fabric. The P–T conditions associated with the final phase of garnet growth were ∼7 kbar and 650 °C. The monazite ages coupled with the phase relations modelled from this multistage garnet indicate that it records two Neoproterozoic tectonothermal events as well as the widespread Ordovician Grampian event associated with Caledonian orogenesis. Thus, this single garnet records much of the Neoproterozoic to Ordovician thermal history in NW Scotland, and highlights the long history of porphyroblast growth that can be revealed by in situ isotopic dating and associated P–T modelling. This approach has the potential to reveal much of the thermal architecture of Neoproterozoic events within the Moine Supergroup, despite intense Caledonian reworking, if suitable textural and mineralogical relationships can be indentified elsewhere.

The Lewisian Gneiss Complex of northwestern Scotland consists of Archaean gneisses, variably reworked during the Proterozoic. It can be divided into three districts – a central granulite-facies district between districts of amphibolite-facies gneiss to the north and south. Recent work has interpreted these districts in terms of separate terranes, initiating a controversy that has implications for how Precambrian rocks are understood worldwide. The northern district of the Lewisian Gneiss Complex (the Rhiconich terrane) is separated from the central district (the Assynt terrane) by a broad ductile shear zone known as the Laxford Shear Zone. This paper reviews the geology of the Laxford Shear Zone, clarifying field relationships and discussing other evidence, to consider whether or not it does indeed represent a terrane boundary. A detailed review of field, geochemical and geochronological evidence supports the recognition of the separate Assynt and Rhiconich terranes. Mafic dykes (the Scourie Dyke Swarm) and granitoids, of Palaeoproterozoic age, occur on both sides of the Laxford Shear Zone and thus the terranes were most probably juxtaposed during the late Archaean to early Palaeoproterozoic Inverian event. Subsequently, the less-competent, more-hydrous amphibolite-facies gneisses of the Rhiconich terrane were affected by later Palaeoproterozoic (Laxfordian) deformation and partial melting, to a greater extent than the more-competent granulite-facies gneisses of the Assynt terrane.

Basement gneiss inliers within the Scottish Caledonides have been conventionally correlated with the Archaean Lewisian Gneiss Complex of the Caledonian foreland. Alternatively, the inliers could represent allochthonous terranes accreted to Laurentia before or during the Caledonian orogeny. Secondary ionization mass spectrometry U–Pb zircon dating indicates that the Ribigill, Borgie, Farr and Western Glenelg basement inliers are characterized by late Archaean protolith ages, and a period of isotopic disturbance in the late Palaeoproterozoic. The data are broadly consistent with correlation between the inliers and components of the Lewisian Gneiss Complex of the Caledonian foreland. The c . 2900 Ma protolith ages support correlation of the Borgie and Farr inliers with the Assynt terrane, and a younger, c . 2800 Ma age for the Ribigill inlier supports correlation with the Rhiconich terrane. None of the studied inliers shows a complete match of protolith and early metamorphic histories with any of the Lewisian basement terranes, but differences between the inliers and the foreland are no greater than those recorded within the foreland basement terranes themselves. Therefore, it remains probable that the dated inlier gneisses formed a distal part of the Laurentian margin prior to final telescoping during the Caledonian orogeny.

U-Pb zircon data from selected gneiss samples provide estimates of the timing of igneous protolith emplacement and granulite facies metamorphism in the Gruinard Bay, Badcall Bay and Lochinver areas of the mainland Lewisian Gneiss Complex of north-west Scotland. Zircons from the gneisses contain complex zoning patterns making them amenable to SHRIMP ion microprobe analysis, with inherited zircon distinguishable from magmatic growth and metamorphic recrystallisation. From Gruinard Bay, a tonalitic granulite yielded a protolith age of 2,825±8 Ma for oscillatory-zoned zircon cores, with recrystallised low-U rims preserving an age of 2,733±12 Ma. An inherited 2,905±15 Ma age component was recognised in a sample of trondhjemitic gneiss, the protolith of which is estimated to have been emplaced at 2,858±11 Ma. Low-U rims yielding an age of 2,729±20 Ma are also present, which together with the rim age from the granulite are considered to date granulite facies metamorphism in the Gruinard region. For comparison with the Gruinard Bay gneisses, a trondhjemitic partial melt and trondhjemite sheet were analysed from the granulite facies rocks at Badcall Bay in the Assynt terrane. Age data do not show definitive relationships with zoning patterns in the zircons, and show a spread along concordia typical of previous ion microprobe age data from the Scourie gneisses. Both samples do, however, show a significant isotopic disturbance at 2,480 Ma, the most recent estimate for the Badcallian metamorphism. Protolith ages for components of the Assynt terrane range between ca. 2,900 and 3,030 Ma, and appear to be older than the 2,860–2,825 Ma gneiss protoliths from the Gruinard Bay area. The timing of granulite facies metamorphism also appears to be different between the two areas, suggesting that the two granulite regions of the mainland may not be related as previously thought. Instead, they are interpreted in terms of a terrane accretion model whereby the two areas evolved as separate blocks prior to their juxtaposition during the Palaeoproterozoic. The boundary between the two terranes is thought to lie to the south of Lochinver, corresponding with the Strathan Line, because of the observed 2,480 Ma age signature to the north and 2,730 Ma signature to the south. By defining a new mainland terrane, the Gruinard Terrane, the number of mainland blocks, which includes the Assynt (central), Rhiconich (northern) and Southern terranes, has been expanded.

The current nomenclature for the Lewisian Gneiss Complex has evolved from lithological and structural correlations made prior to any dating. Initial (flawed) geochronological studies gave some names an apparent chronological standing but, as work advanced, fitting events into a coherent regional framework became increasingly difficult. Modern dating studies have shown that the Lewisian Gneiss Complex was progressively assembled from disparate blocks of Archaean continental crust and juvenile Proterozoic arcs, which satisfy the definition of terranes. Each terrane had its own separate accretionary and metamorphic history followed by a common history once juxtaposed against other terranes. Based on a new compilation of modern geochronology allied to the many detailed structural and metamorphic studies, this paper proposes a new systematic terminology for the Lewisian Gneiss Complex that is more applicable to this new tectonic framework.

In the south of the Nuuk region of West Greenland our 1980s mapping recognized four Archaean gneiss terranes (Fœringehavn, Tre Brodre, Tasiusarsuaq and Akia terranes) with different protolith ages and separate early tectonothermal histories. Later in the Archaean these were juxtaposed and then experienced the same 2700–2500 Ma tectonothermal events. Here we abandon extrapolation of only these four terranes across the whole region, and distinguish two new terranes in the NE. The northernmost Isukasia terrane (previously regarded as the northernmost exposure of the Fœringehavn terrane) consists of Palaeoarchaean rocks (>3600 Ma) tectonically bounded on its south by the 3075–2960 Ma Kapisilik terrane; these were juxtaposed and metamorphosed together by 2950 Ma. The previously recognized Fœringehavn terrane to the SW is another, separate entity of Palaeoarchaean rocks that was juxtaposed with adjacent terranes only after c . 2800 Ma. Hence in an increasingly complex regional model, there were several mid- to Neoarchaean terrane assembly events, with superimposed ‘orogenies’ from c . 2950 Ma until after 2700 Ma. Although the Fœringehavn and Isukasia terranes were incorporated into the later Archaean terrane collage at different times, they might be fragments from a larger Palaeoarchaean complex rifted apart from c . 3500 Ma onwards.

40 Ar/ 39 Ar and Rb–Sr mineral ages have been determined from various lithologies exposed in the Caledonian foreland and structurally overlying thrust nappes of north Sutherland, Scotland. Rb–Sr muscovite ages of c . 428, c . 421 and c . 413 Ma obtained from Moine Thrust Zone mylonites are interpreted to date closely regional thrusting during the Late Silurian to Early Devonian. 40 Ar/ 39 Ar muscovite ages within the lower parts of the Moine nappe are mostly anomalously old with respect to Rb–Sr analyses of muscovites from the same samples; it is likely that this discrepancy results from a component of extraneous or ‘excess’ argon. 40 Ar/ 39 Ar hornblende ages and Rb–Sr and 40 Ar/ 39 Ar muscovite ages obtained from structurally higher metamorphic units in the Caledonian thrust nappes generally range between c . 440 Ma and c . 410 Ma. These ages are interpreted to date cooling during and following ‘D 2 ’ regional thrusting and folding within internal sectors of the nappe sequence. A possible tectonic model involves the Silurian collision of Baltica with Scottish segments of Laurentia resulting in the Scandian orogeny and broadly coeval Moine Thrust Zone. D 2 structures were superimposed on structures and metamorphic fabrics formed during a regional Mid-Ordovician tectonothermal event dated previously at c . 470–460 Ma. Syn-D 2 temperatures were generally >600°C and sufficient to achieve more or less complete thermal rejuvenation of Rb–Sr and 40 Ar/ 39 Ar isotopic systems in muscovite and hornblende, even in areas of low D 2 strain.

The West Highland granite gneiss suite in Inverness-shire, Scotland, represents a series of S-type, anatectic granites formed by partial melting of host Neoproterozoic metasediments of the Moine Supergroup. U–Pb (SHRIMP) dating of zircons from a member of the suite, the Fort Augustus granite gneiss, indicates that the granitic protolith to the gneiss was intruded at 870±30 Ma. This is indistinguishable from the published age determined by the same method for the Ardgour granite gneiss at Glenfinnan, thus supporting the assumption that the various members of the West Highland granite gneiss are part of a single intrusive suite. The spread of ages from the zircon cores (1626–947 Ma) is interpreted to indicate a Proterozoic source terrain for the Moine sediments that were later melted to form the granitic protolith. A U–Pb age of 470±2 Ma obtained for titanite in the Fort Augustus granite gneiss is interpreted to date amphibolite-facies metamorphism during the early to mid-Ordovician Grampian Orogeny. The emerging similarity in the timing of this event either side of the Great Glen Fault implies that this structure does not juxtapose crustal blocks with significantly different histories with respect to the Grampian Orogeny.

Graham Park writes: The terrane model of Kinny et al. (2005; see also Friend & Kinny 2001) presents a radically different picture of the Lewisian complex from the ‘traditional’ view of a single piece of Archaean crust that has been subsequently modified and augmented in various Proterozoic events. The authors are to be congratulated on their attempt to rationalize a very confused terminology that has arisen over many decades, and for forcing us to rethink our views. I have reservations about their re-definition of ‘Laxfordian’, but my comments mainly concern the validity of the terrane model as applied to the Lewisian and suggestions to refine it (see Park 2005). I agree that uncritical use of terms such as ‘Scourian’, ‘Laxfordian’ and ‘Scourie dyke suite’ can obscure genuine differences of rock, process or event (e.g. see Park et al . 2002). Most of the authors’ suggested revisions seem sensible, with the exception of their proposed restriction of ‘Laxfordian’ to a c. 1.74 Ga metamorphic event at Loch Laxford. I believe that this term is more useful in its current sense as a general name for the series of events etc. that affected the entire Lewisian complex after the incorporation of the early Palaeoproterozoic material (e.g. the Loch Maree Group and the ‘Scourie’ dykes). I have retained the traditional usage below for convenience. According to this concept, distinction is made between ‘suspect terranes’ and ‘displaced terranes’ i.e. those that are exotic or allochthonous. In order to qualify as a displaced terrane, a piece of crust should display evidence of significant displacement relative to its neighbour. Differences in geological history between two adjacent pieces of crust may not in themselves be sufficient evidence of large relative displacements between them. Transverse or oblique convergence between two continental plates inevitably involves the subduction …

Within amphibolite facies Peninsular gneisses in the south of the Dharwar craton, units of Sargur supracrustal rocks contain ultrabasic enclaves. One of these enclaves is an orthopyroxenite which comprises bronzite, spinel and minor phlogopite preserving coarse-grained, relic textures of probable igneous origin. After incorporation into the gneisses the enclave evolved through several distinct stages, elucidation of which allow an assessment of its metamorphic history.Firstly, deformation during closed system, anhydrous recrystallisation caused the coarse-grained textures to be partially overprinted by similar mineral assemblages but with a granoblastic texture. Secondly, open system hydration caused retrogression of the bronzite to alumino-gedrite at the margins of the enclave. Subsequently, the penetration of these fluids along grain boundaries caused reactions between spinel and bronzite to produce reaction pockets carrying assemblages of peraluminous sapphirine associated with cordierite and talc. The differences in the mineral assemblages in each pocket coupled with slight variations in their chemistry, suggest that equilibrium did not develop over the outcrop. Because sapphirine + magnesite is present in some pockets, it is evident that CO2 was also a component of the fluid.Phase relations from the MASH portion of the FMASH system, to which the chemistry of the reaction pockets approximates, suggest that the hydrous metamorphism causing the changes depended upon the assemblage enstatite + spinel + vapour which exists at PT conditions above the position of I16, ∼760°C at 3 kbar and below I21 at ∼765°C at 5.6 kbar (Seifert, 1974, 1975), where sapphirine is replaced by kornerupine. The suggested path of reaction occurred between I18 and I21. Subsequent reactions related to I20 cause the formation of cordierite. Talc formation has to be modelled in a different reaction grid.The metamorphism recorded by these reactions is thus at a maximum of amphibolite facies and is interpreted to have formed during the uplift and cooling history of the gneiss complex when hydrous fluids were free to migrate. Given the complex high-grade metamorphic history of this part of the Dharwar craton this event is likely to be late Archaean or Palaeoproterozoic in age.

Musgravite, Be(MgFeZn)2Al6O12, is associated with norbergite and minor chlorite in a Precambrian calcite marble within the gneissic basement in an internal part of the Caledonian mobile belt in Dove Bugt, North-East Greenland. It commonly occurs as vitreous black, idioblastic crystals (2O368.74-70.63; Tot, Fe as FeO 6.76-7.89; MgO 12.17-13.98; and ZnO 3.22-4.47. ICP-MS analysis also revealed significant trace amounts of V 249 ppm, Cr 740 ppm and Ga 178 ppm. The crystallographic parameters and composition are broadly in accord with those of musgravite from the two other recorded occurrences, in Precambrian high-grade terrains in Australia and Antarctica, although mineral and rock associations in these localities differ from those in Dove Bugt. The source of beryllium in each of the musgravite occurrences is uncertain, although a metasomatic source related to granite emplacement is favoured for the occurrence in North-East Greenland.

The arrested, prograde amphibolite- to granulite-facies transition at Kabbaldurga, south India, overprints Archacan amphibolite-facies nebulitic gneisses and the late Archaean Closepet granite. Previous studies have shown that this facies transition was controlled by a channelled influx of a dehydrating fluid, assumed to be CO2, at ∼750°C and 5.5 kbar confining pressure. The effect of this type of prograde transition on zircon U−Pb isotopic systematics and whole-rock geochemistry has been studied using ∼1 kg amphibolite-facies, transitional and granulite-facies domains from a single block of gneiss. The zircon populations from all three domains have essentially similar morphology and U−Pb systematics. This similarity shows that at the conditions under which the prograde granulite-facies transition took place via fluid influx, the zircon U−Pb systematics were not disturbed by the process. Using the pooled data from all three domains, it is concluded that the protolith of the gneiss formed at 2965±4 Ma (2σ), and that zircons also grew during an anatectic event common to all domains at 2528±5 Ma. The granulite-facies metamorphism has not been dated directly due to the lack of response to the zircon U−Pb isotopic systematies to it. However, field and petrographic criteria dictate that its maximum age is 2528±5 Ma, the age of the anatectic event common to each domain in the gneiss block, which was overprinted during the granulite-facies event. For most major and trace elements, consistent enrichment or depletion trends associated with the transition to granulite facies cannot be identified with confidence. However, the granulite-facies portion is LREE (light-rare-earth-element)-enriched and H (heavy) REE-depleted compared with the amphibolite-facies domain, and the transitional domain is at intermediate values. The isotopic and geochemical evidence presented supports the conclusion that the granulite-facies charnockitic rocks at Kabbaldurga were not formed by removal of an anatectic melt, but that they formed later by simple metamorphic overprint of amphibolite-facies rocks.

The Tasiusarsuaq terrane in the Fiskenaesset-Nuuk region of southern West Greenland is interpreted as a late Archean tilted crustal section modified by deformation. The northern part of the terrane was affected by granulite-facies metamorphism at about 2810 Ma, but large areas were subsequently retrogressed to amphibolite facies. In the prograde amphibolite- to granulite-facies transition in inner Bjornesund a relatively sharp, but highly irregular front that crosses structures and lithological boundaries on outcrop scale separates orthopyroxene-free from orthopyroxene-bearing rocks. Outcrop relations resemble in some ways examples of "arrested charnockite formation" described from high-grade regions in southern India and Sri Lanka. Relations in the transition zone in Bjornesund appear to be the result of late, local influx of a fluid, presumably rich in $$CO_{2}$$, that permitted orthopyroxene to grow at the expense of hornblende. Regional considerations suggest a genetic connection between granulite-faci...

In the Godthabsfjord region of southern West Greenland a NE-SW-trending belt of rocks of very varied age and origin, here named the Akulleq terrane, is separated by major faults from more extensive blocks of typical high-grade Archaean rocks that, although they are superficially similar, have different ages and metamorphic histories. The continental crust that forms the block to the north-west, the Akia terrane, was accreted between ea. 3200 and 2980 Ma, and that forming the block to the south-east, the Tasiusarsuaq terrane, between 2920 and 2800 Ma. It is suggested that the Godthabsfjord belt is the result of collision of the two continental blocks between 2800 and 2650 Ma. The rocks of the Akulleq terrane are interpreted as fragments of different parts of the crust that originally separated the two continents. They include early Archaean continental crust, possible oceanic crust, and acid to intermediate rocks of intrusive and possibly also extrusive origin that may have been generated in a subduction-related environment.

Recent work has provided a new tectono-metamorphic framework for the Archaean evolution of the Godthabsfjord region which can be divided into discrete tectonic units each characterized by different magmatic, structural and metamorphic histories. These tectonic units were assembled in the late Archaean, post-2750 Ma. The present structural patterns seen in the region date from the period 2750–2600 Ma and do not reflect earlier structural events, as was previously considered. Late Archaean structural evolution began with a period of interthrusting of early Archaean sialic and late Archaean oceanic crustal segments coeval with, or followed by, isoclinal nappe formation under amphibolite-facies conditions. Linear fabrics in early Archaean rocks were rotated into parallelism with those newly formed parallel to F1 hinge lines in the late Archaean rocks. A second major thrusting event emplaced granulite facies rocks structurally on top of the amphibolite facies rocks. Subsequent to this event a series of F2 upright or steeply inclined basin and dome structures formed, some of which have sheath morphology. This second folding episode acting upon the isoclinal nappes produced the regional scale interference patterns observed. The early linear fabrics were re-oriented around these second generation structures. Data are produced from selected areas to illustrate the fabric patterns and structural continuity that exists throughout the region.

On the Nuussuaq peninsula, Western Greenland sedimentary deposits of glass spherules also contain high Ir, Co, Ni, and Cu anomalies. The iron-rich silicate glass spherules (to ∼3 wt% NiO, ∼35 wt% FeO) are highly circular in cross section. They show surface dissolution, smectite replacement and calcite infilling of vesicles, though many glasses are optically unaltered. They are strikingly heterogeneous, with schlieren outlining counter-flowing convection cells. Their pronounced Fe-Ni correlation is unlike volcanic suites, but is explained by mixing between basaltic melt and an enriched iron-nickel source. Distinctive nickel-spinel (∼7-10 wt% NiO) contains very nickel-rich cores. Occasional glass spherules show compositional gradients toward resorbed silicates, (plagioclase, clinopyroxene); isotropic plagioclase has anomalous texture comparable to impact-melted lunar breccias. Their anomalously high copper and sulfur (to ∼1%) have lead to an explanation as products of fire-fountaining of exotic or picritic Disko lavas; they would be perhaps the only non-impact occurrence of Ni spinel. Since their discovery, better criteria for recognition of spherules ejected from large impacts have been established, and greater variations in meteorite chemistry as potential projectiles have been described. New mineralogical and petrographic textural data for the Nuussuaq spherules suggest they should be reinterpreted as impact ejecta; the highly oxidized Ni-spinel is a very characteristic signature of meteorite impact ejecta. Delicate preservation features rule out substantial sedimentary reworking, and spherule bed thicknesses imply a large source crater. Stratigraphically, the spherule beds are poorly constrained, but nannofossils and magnetostratigraphy place them close to onset of the West Greenland flood lavas (ca. 61-62 Ma). They share many characteristics with massive native iron localities in dykes and lavas up to >100 km away on Disko (Qeqertarsuaq) Island, but their precise relationship remains to be established.

Since the pioneering days of Peach, Horne and Clough the Lewisian Gneiss Complex has become a classic geological area, representing the first piece of gneissic crust recognized as preserving a section through lower crust. Because of its accessibility it has become a place where new hypotheses have been tried and tested with varying degrees of success and considerable controversy. Whilst there have been copious papers on the detailed geology and many summaries in the form of regional guides and book chapters, a series of modern memoirs as a data base on the Lewisian has been lacking. The lack of a series that acted as a data repository has probably been detrimental to advances in understanding the evolution of this complex piece of geology. This memoir from the Geological Society at least addresses this gap for the Gairloch area with a meticulous compilation and description of the geological information by Park. The production of this volume represents the culmination of a lifetime’s work by one who has been involved with unravelling the geology of the Gairloch area for about forty years. The detailed knowledge of outcrops and their relations probably cannot be bettered. The Gairloch area has long been proposed as the critical area for the interpretation of the Lewisian Gneiss Complex. Whilst the arguments over how important the area actually is will probably continue, it is certainly very important to the understanding …

Since the early work of Jehu and Craig (1927) the metasedimentary gneisses of the Leverburgh belt and the intrusive plutonic rocks of the South Harris igneous complex have been the subject of many geological studies (for example, Dearnley 1963; Coward et al. 1969; Wood 1975; Graham 1980). The establishment of the age of the South Harris igneous complex anorthosite component at 2180 ± 60 Ma, and an age of 1870 ± 30 Ma for a subsequent granulite facies event (Cliff et al. 1983) has helped to resolve two of the debates over the geological evolution of the area. First, the correlation with the Scourie dykes of basic bodies cutting the complex (Dearnley 1963; Witty 1975; Dickinson and Watson 1976; Heyes 1978; Horsley 1978). Second, the debate over the timing and number of granulite facies events. Dearnley (1963) suggests that evidence for an early granulite facies event (Scourian) was preserved in the pyroxene granulite (norite) bordering the Sound of Harris (Fig. 1). However, Palmer (1971) and Graham (1980) consider that the norite forms a part of the South Harris igneous complex, in which case the granulite facies assemblages would thus be Laxfordian. An important component of the Leverburgh belt are units comprising highly aluminous quartz-rich metasedimentary rocks with variably retrogressed high-pressure granulite-facies mineral assemblages (Dearnley 1963; Coward et al. 1969; Wood 1975; Dickinson and Watson 1976). These lithologies in particular demonstrate the severe effects of the Laxfordian remobilisation and deformation. Relict structures, which might indicate the pre-Laxfordian history of these rocks, . . .

Granites and granite pegmatites composing the ∼ 2550-Ma Qorqut granite complex occur in a SSW-NNE trending linear belt >150 km long extending through the Buksefjorden-Ameralik-Godthabsfjord region of southern West Greenland. The main body of the complex crops out over a distance of 50 km from Ameralik to Kapisigdlit kangerdluat and reaches a maximum outcrop width of 18 km between Storo and Qorqut. Around Qorqut the complex comprises three main groups of granites: early leucocratic granites, various grey biotite granites, and late aplogranite-granite pegmatites. Within the 1500-m vertical section available in this area the complex has a tripartite structure comprising a lower zone dominantly of polyphase granite, an intermediate zone where country rock occurs as rafts in polyphase granite with a complex sheeted structure, and an upper zone dominantly of country rock sheeted by granite. Fifty-two specimens of granite have been analyzed for major, minor, and some trace elements. Geochemical variation within the complex is consistent with either fractional crystallization or partial melting, but in both cases, feldspar + biotite must have been involved either as fractionating phases or as residual phases during melting to account for the trace element chemistry. Two possible models for the generation of the complex are either anatexis of granulite facies rocks in the lower crust following an influx of volatiles and heat from the mantle or melting at intermediate depths of amphibolite facies rocks with volatiles supplied by breakdown of hydrous phases.

Gabbroic-textured shonkinitic rocks of unknown age cut deformed Archaean dunites and peridotites in the area immediately north of Salem. Magnesite formation post-dates their intrusion. The location of these rocks and the magnesite is thought to be controlled by the intersection of two major basement lineaments. A sequence of crystallization of the minerals and rocks from early undersaturated to late oversaturated rocks is established. Compositions of the coexisting mineral phases are reported here for the first time. Clinopyroxenes, which are relatively unevolved, commonly coexist with olivine (chrysolite) and sanidine. Bronzite, often intergrown with magnetite and green spinel, is occasionally present. In the most undersaturated rocks nepheline-sanidine fingerprint intergrowths occur, whilst perthitic sanidine, albitic plagioclase, and quartz are present in over-saturated rocks. Late in the crystallization sequence, in over-saturated rocks, PH2O appears to have increased allowing the formation of amphibole as well as clinopyroxene.

Chemically complex pyroxenes which occur in early Proterozoic tholeiitic dolerite dykes in southern West Greenland have been investigated using back-scattered electron (BSE) imagery, X-ray mapping and electron microprobe analysis. A wide variety of compositions occur within individual pyroxene grains in these rocks. They can be explained by simultaneous nucleation of different pyroxenes, the evolution of domains around these nucleii as a response to differential chemical gradients and the sequential precipitation of different pyroxenes at progressively lower temperatures. As an example, the individual grains in one dyke sample contain domains of bronzite, hypersthene, magnesium pigeonite, augite, and subcalcic augite. Olivine in this sample varies in composition from Fo70 to Fo33, although individual grains are only weakly zoned. The wide variation in pyroxene and olivine compositions suggests ranges of crystallization temperatures from c.1250° to as low as 825°C. Such compositionally variable pyroxenes are possibly characteristic of hypabyssal tholeiitic rocks.

The Faeringehavn–Tre Brodre area consists of three distinct terranes tectonically jux-taposed by a previously unrecognized event. Contacts between the terranes are mylonitic shear zones truncating lithological units in adjacent terranes. The terrenes are: (1) Faeringehavn terrane, largely composed of early-Archaean Amitsoq gneiss cut by younger Archaean granitic gneiss defined here as the Satut gneiss; (2) Tre Brodre terrane comprising mid-Archaean Malene supracrustal rocks, anorthosite complex and polyphase Nuk gneisses; (3) Tasiusarsuaq terrane largely comprising mid-Archaean Nuk gneisses affected by c. 2800 Ma granulite facies metamorphism. The Tasiusarsuaq terrane is structurally above both the Faeringehavn and Tre Brodre terranes which at c. 2800 Ma experienced a lower grade of metamorphism. Juxtaposition of the terranes took place between 2800 and 2500 Ma and involved thrusting and crustal thickening. Subsequent re-equilibration involved folding, steeply inclined shear belts, intrusion of synkinematic granitoids under amphibolite facies conditions and retrogression of granulite facies assemblages. This thrusting post-dates the 2800 Ma granulite facies metamorphism. It is younger and distinct from the thrusting postulated to explain the intercalation of the early-Archaean Amitsoq gneisses and the mid-Archaean Malene supracrustal rocks, associated with intrusion of the Nuk gneiss precursors described from Godthabsfjord. The tectonic breaks provide, for the first time, marker horizons which can be used to assess the amount and type of late Archaean deformation in the southern Godthabsfjord region.

The investigation of Proterozoic basic dykes in southern West Greenland stemmed from the programme of systematic mapping of the Archaean craton in that region by the Geological Survey of Greenland (GGU). This work began in the southern Frederikshåb region in the early 1960s (Jensen, 1968, 1969) and progressed northwards, from bases in the Fiskenæsset (Kalsbeek & Myers, 1973; GGU, 1976), Godthåb (Allaart et al., 1977) and Sukkertoppen areas (Allaart et al., 1978). The results of most of this mapping work were summarized by Bridgwater et al. (1976) and compiled onto a 1:500 000 scale geological map sheet by Allaart (1982). The distribution of the major Proterozoic dykes which cut the entire region is shown on this map. While the basic dykes are individually minor intrusions, many are up to 50 metres wide and continuous for several tens of kilometres, and collectively they represent a major magmatic event. As many of the Archaean terrains of the world possess Proterozoic basic dyke swarms, their compositions are crucial to a correlation of events from one craton to another and to an understanding of crustal and mantle evolution after the world-wide late Archaean sialic crust-forming event.

Cr-rich Mg-sapphirines, which are more aluminous in composition than R2+:R3+:Si of 7:9:3 and are therefore considered to be peraluminous, are shown to have an inverse linear relationship between octahedrally co-ordinated Cr and Al. This balanced substitution, which has not previously been described, may be used to explain the major chemical variations exhibited in the described minerals. Since the peraluminous Cr-rich sapphirines are idioblastic it is suggested that they are stable members of the assemblages within which they occur. Peraluminous sapphirine is usually considered to be metastable but in this instance it is possible that chromium may assist in stabilizing its development.

The discovery, at Kabbaldurga quarry, Karnataka, south India1–3, of patches of charnockite apparently in an arrested state of development has refocused attention on the mechanism of charnockite formation. In particular, the relative roles of CO2 and H2O during high grade metamorphism and charnockite development can be studied in such an area. During an influx of a CO2-rich volatile phase, H2O may be liberated as a result of the reduction in the state of hydration by the breakdown of hydrous minerals4. The P–T conditions in which this occurs are important because, with an increase in XH2O, conditions which may allow anatexis could be initiated in advance of such a CO2-rich phase. Indeed, Weaver5 has suggested that the K-rich acid charnockites found at Pallavaram, Madras6, developed due to changes in volatile composition before the onset of charnockite conditions. The exposure at Kabbaldurga quarry is interpreted here as displaying evidence that such a process has taken place.

Information regarding the processes whereby granulite facies rocks were formed or destroyed may be gained from a study of the morphology of transitional zones. Both transitions have a complex 3D structure which may include anatectic rocks suggesting that melting can play an important role. Because of the intimate relationship of granulite and amphibolite facies assemblages such transitions appear to have been largely fluid controlled. The nature of this control could be a simple compositional change or a more complicated mechanism. Some transitions seem to have had a flux of dehydrating fluids which lead to melting in amphibolite facies rocks, whilst in others melting in the granulite facies rocks occurred.

Gneisses affected by ca. 2800 Ma granulite facies metamorphism outcrop at Kangimut sammisoq, Ameralik fjord, southern West Greenland. At ca. 2700 Ma these granulite facies rocks were thrust over a terrane dominated by >3600 Ma AmItsoq gneisses and a terrane dominated by 2800–2750 Ma Ikkattoq gneisses, which had not undergone the granulite facies metamorphism. Following the thrusting event, further deformation gave rise to folds and subvertical shear zones formed between ca. 2700 and 2550 Ma. Thrusting and subsequent folding and shearing were accompanied by retrogression of the granulite facies rocks under amphibolite facies conditions. From Nd, Sr, Hf and U-Pb isotopic systematics the rocks at Kangimut sammisoq have an age of no more than 3000 Ma. However, they contain variable amounts of unradiogenic Pb derived from early Archaean rocks such as >3600 Ma AmItsoq gneisses. Retrogression involved an increase in the Pb content of the gneisses at Kangimut sammisoq. Some or all of this Pb was introduced at the time of retrogression via hydrous fluids that were derived from, or passed through, the underlying tectonic units of AmItsoq and Ikkattoq gneisses.