A crystal-rich volcaniclastic sandstone in the lower Peltura scarabaeoides Zone at Ogof-ddû near Criccieth, North Wales, yields a U–Pb zircon age of 491∓1 Ma. This late Late Cambrian date indicates a remarkably young age for the Cambrian–Ordovician boundary whose age must be less than 491 Ma. Hence the revised duration of the post-Placentian (trilobite-bearing) Cambrian indicates that local trilobite zonations allow a biostratigraphic resolution comparable to that provided by Ordovician graptolites and Mesozoic ammonites.

Middle Triassic radiolarian bedded cherts in the Mino Belt, central Japan, include a sequence showing an abrupt facies change from the lower to the upper, where grey–black bedded cherts enriched in carbonaceous matter and framboidal pyrite are overlain by brick-red hematitic bedded cherts. Brownish-yellow chert enriched in goethite and purple-red chert occur at the boundary between the grey–black bedded cherts and the brick-red bedded cherts. This facies change is in accordance with stratigraphic variations of geochemical characteristics; the lower section grey–black bedded cherts, compared with the upper section brick-red bedded cherts, are enriched in Ctot and Stot, and are characterized by lower MnO/TiO2, higher FeO/Fe2O3* (total iron as Fe2O3) and more variable Fe2O3*/TiO2 values. Some of the lower section samples, in addition, are characterized by an enrichment in some transition metals (Ni, Cu, and Zn). The covariation of mineralogical and geochemical characteristics indicates that sedimentary environments and diagenetic processes were different between the lower and the upper section bedded cherts. During the deposition of the lower section bedded cherts, the sedimentary environment was anoxic and bacterial sulphate reduction occurred during the early diagenetic stage. In contrast, the upper section bedded cherts were subjected to less reducing diagenetic processes; active sulphate reduction did not occur. The change of sedimentary environment and diagenetic process at the site of deposition is likely to be attributed to the fluctuated concentration of dissolved oxygen in the water mass of a semi-closed marginal ocean basin, which was potentially caused by sea-level change that occurred during Middle Triassic time.

A large number of Lower Silurian (Llandovery) K-bentonite beds have been recorded from northwestern Europe, particularly in Baltoscandia and the British Isles, but previous attempts to trace single beds regionally have yielded inconclusive results. The present study suggests that based on its unusual thickness, stratigraphic position and trace element geochemistry, one Telychian ash bed, the Osmundsberg K-bentonite, can be recognized at many localities in Estonia, Sweden and Norway and probably also in Scotland and Northern Ireland. This bed, which is up to 115 cm thick, is in the lower–middle turriculatus Zone. The stratigraphic position, thickness variation and geographic distribution of the Osmundsberg K-bentonite are illustrated by means of 12 selected Llandovery successions in Sweden, Estonia, Norway, Denmark, Scotland and Northern Ireland. In Baltoscandia, the Osmundsberg K-bentonite shows a trend of general thickness increase in a western direction suggesting that its source area was located in the northern Iapetus region between Baltica and Laurentia. Because large-magnitude ash falls like the one that produced the Osmundsberg K-bentonite last at most a few weeks, such an ash bed may be used as a unique time-plane for a variety of regional geological and palaeontological studies.

At the submerged margins of the North Atlantic, andesitic to dacitic and basaltic volcanic rocks occur together. The silicic rocks were derived by processes requiring the presence of continental crust (crustal anatexis and/or contamination of mafic magmas) while the majority of the basaltic lavas had little or no contact with continental crust. We report 40Ar–39Ar incremental heating ages for several dacitic and basaltic rocks recovered from three offshore localities of the North Atlantic Igneous Province. Dacitic lavas and tuffs at the southeast Greenland margin and trachytic lavas in the Scottish Hebrides erupted contemporaneously with basaltic lavas at 62–61 Ma. In contrast, the silicic lavas from the northern Rockall Trough (offshore western Scotland) and the Vøring Plateau (offshore Norway) erupted at ∼55 Ma followed shortly by basaltic volcanism. At this time, silicic magmatism at the southeast Greenland margin had ceased and only oceanic basalts were erupted. Similarly, ∼55 Ma lavas on the southwest Rockall Plateau are wholly basaltic. The compositions of all of the dated silicic volcanic rocks are consistent with derivation from partial melting of either continental crust or sediments. The heat necessary for partial melting appears to have been provided by basaltic magmas. Therefore, the existence of the silicic rocks indicates the presence of continental crust as well as a stable tectonic environment that allowed the stagnation and pooling of basaltic melts within the crust. With this in mind, it is apparent that at 62–60 Ma, both western and eastern sides of the present North Atlantic margins were characterized by extensional environments within continental crust that were restrictive to the passage of mafic magmas. By 55 Ma, at the time of continental breakup, the proximal margins at southeast Greenland and the Rockall Plateau were devoid of continental crust. But the presence of 55 Ma silicic magmatism on the eastern North Atlantic margin can be attributed to a broader zone of magmatism and sediment-filled Mesozoic rift basins.

Global sea-level fluctuated markedly during the early Silurian, probably as a result of the waxing and waning of ice-sheets in the South American portion of Gondwana. The highest sea-levels of the Silurian are recorded by the Telychian upper crispus–lower griestoniensis and spiralis–lower lapworthi biozones. Other highstands occurred in the early Aeronian, during the convolutus Zone (mid Aeronian), guerichi Zone and late turriculatus Zone (early Telychian), and early Sheinwoodian. Low sea-levels characterized much of the argenteus and sedgwickii zones (Aeronian), the utilis Subzone (late guerichi–early turriculatus zones, early Telychian), the late Telychian (commencing in the mid lapworthi Zone) and, after a period of apparently only small amplitude sea-level fluctuations in the late Sheinwoodian and earliest Homerian, the mid–late Homerian, in particular the early nassa Zone. Facies (and faunal) changes in the Lower Silurian do not support the P and S model of Jeppsson and others, but are consistent with the sea-level changes proposed herein. Mid Telychian marine red beds appear to have been deposited during a minor sea-level fall immediately after a period of very high sea-levels, rather than during a transgressive episode as previously suggested. Comparison of the sea-level curve presented herein with those constructed in the past is hampered by the lack of precision currently possible in the correlation of early Silurian deep water (graptolitic) and shallow water (shelly) sequences. Improving the precision of this correlation should be a priority for future research.

The Canarian Archipelago is a group of volcanic islands on a slow-moving oceanic plate, close to a continental margin. The origins of the archipelago are controversial: a hotspot or mantle plume, a zone of lithospheric deformation, a region of compressional block-faulting or a rupture propagating westwards from the active Atlas Mountains fold belt have been proposed by different authors. However, comparison of the Canarian Archipelago with the prototypical hotspot-related island group, the Hawaiian Archipelago, reveals that the differences between the two are not as great as had previously been supposed on the basis of older data. Quaternary igneous activity in the Canaries is concentrated at the western end of the archipelago, close to the present-day location of the inferred hotspot. This is the same relationship as seen in the Hawaiian and Cape Verde islands. The latter archipelago, associated with a well-defined but slow-moving mantle plume, shows anomalies in a plot of island age against distance which are comparable to those seen in the Canary Islands: these anomalies cannot therefore be used to argue against a hotspot origin for the Canaries. Individual islands in both archipelagoes are characterized by initial rapid growth (the ‘shield-building’ stages of activity), followed by a period of quiescence and deep erosion (erosion gap) which in turn is followed by a ‘post-erosional’ stage of activity. The absence of post-shield stage subsidence in the Canaries is in marked contrast with the major subsidence experienced by the Hawaiian Islands, but is comparable with the lack of subsidence evident in other island groups at slow-moving hotspots, such as the Cape Verdes. Comparison of the structure and structural evolution of the Canary Islands with other oceanic islands such as Hawaii and Réunion reveals many similarities. These include the development of triple (‘Mercedes Star’) rift zones and the occurrence of giant lateral collapses on the flanks of these rift zones. The apparent absence of these features in the post-erosional islands may in part be a result of their greater age and deeper erosion, which has removed much of the evidence for their early volcanic architecture. We conclude that the many similarities between the Canary Islands and island groups whose hotspot origins are undisputed show that the Canaries have been produced in the same way.

The Lower Silurian Osmundsberg K-bentonite is a widespread ash bed that occurs throughout Baltoscandia and parts of northern Europe. This paper describes its characteristics at its type locality in the Province of Dalarna, Sweden. It contains mineralogical and chemical characteristics that permit its regional correlation in sections elsewhere in Sweden as well as Norway, Estonia, Denmark and Great Britain. The <2 μm clay fraction of the Osmundsberg bed contains abundant kaolinite in addition to randomly ordered (RO) illite/smectite (I/S). Modelling of the X-ray diffraction tracings showed the I/S consists of 18% illite and 82% smectite. The high smectite and kaolinite content is indicative of a history with minimal burial temperatures. Analytical data from both pristine melt inclusions in primary quartz grains as well as whole rock samples can be used to constrain both the parental magma composition and the probable tectonic setting of the source volcanoes. The parental ash was dacitic to rhyolitic in composition and originated in a tectonically active collision margin setting.

Whole rock chemical fingerprinting of coeval beds elsewhere in Baltoscandia produced a pronounced clustering of these samples in the Osmundsberg field of the discriminant analysis diagram. This, together with well-constrained biostratigraphic and lithostratigraphic data, provides the basis for regional correlation and supports the conclusion that the Osmundsberg K-bentonite is one of the most extensive fallout ash beds in the early Phanerozoic. The source volcano probably lay to the west of Baltica as part of the subduction complex associated with the closure of Iapetus.

Explosive volcanism has dominated the large phonolitic shield volcano of Tenerife, the Las Cañadas edifice, for the last 1.5 m.y. Pyroclastic deposits of the Bandas del Sur Formation are exposed along the southern flanks, and record the last two of at least three long-term cycles of caldera-forming explosive eruptions. Each cycle began with flank fissure eruptions of alkali basalt lava, followed by minor eruptions of basanite to phonotephrite lavas. Minor phonotephritic to phonolitic lava effusions also occurred on the flanks of the edifice during the latter stages of the second explosive cycle. Non-welded plinian fall deposits and ignimbrites are the dominant explosive products preserved on the southern flanks. Of these, a significant volume has been dispersed offshore. Many pyroclastic units of the second explosive cycle exhibit compositional zonation. Banded pumice occurs in most units of the third (youngest) explosive cycle, and ignimbrites typically contain mixed phenocryst assemblages, indicating the role of magma mixing/mingling prior to eruption. At least four major eruptions of the third cycle began with phreatomagmatic activity, producing lithic-poor, accretionary lapilli-bearing fallout and/or surge deposits. The repeated, brief phase of phreatomagmatism at the onset of these eruptions is interpreted as reflecting an exhaustive water supply, probably a small caldera lake that was periodically established during the third cycle. Accidental syenite becomes an increasingly important lithic clast type in ignimbrites up-sequence, and is interpreted as recording the progressive development of a plutonic complex beneath the summit caldera.

Successive eruptions during each explosive cycle increased in volume, with the largest eruption occurring at the end of the cycle. More than ten major explosive eruptions vented moderately large volumes (1−[ges ]10 km3) of phonolitic magma during the last two cycles. Culminating each explosive cycle was the emplacement of relatively large volume (>5−10 km3) ignimbrites with coarse, vent-derived lithic breccias, interpreted to record a major phase of caldera collapse. In the extracaldera record, explosive cycles are separated by ∼0.2 m.y. periods of non-explosive activity. Repose periods were characterized by erosion, remobilization of pyroclastic deposits by discharge events, and pedogenesis. The current period of non-explosive activity is characterized by the construction of the Teide-Pico Viejo stratovolcanic complex within the summit caldera. This suggests that eruptive hiatuses in the extracaldera record may reflect effusive activity and stratovolcano or shield-building phases within the summit caldera. Alternating effusive and explosive cycles have thus been important in the volcanic evolution of the Las Cañadas edifice.

The Lilljeborgfjellet Conglomerate Formation composes the lower part of the alluvial Siktefjellet Group of northwestern Spitsbergen's Old Red Sandstone succession. Siktefjellet strata are of late Silurian or early Devonian age, but lack precise age-diagnostic fossils. They are unconformably overlain by conglomerates and sandstones of the Red Bay Group, which contain a well established fish fauna of Lochkovian age. The Lilljeborgfjellet Conglomerate rests with a major unconformity on high-grade (with eclogites) schists and gneisses, with associated corona gabbros and granitic gneisses. Previous isotope-age studies have shown that these igneous rocks yield U/Pb ages of c. 950 Ma, and that the eclogite facies metamorphism may be of Caledonian or late Neoproterozoic age. The high P/high T rocks are intercalated with and overlain by schists affected only by Caledonian amphibolite facies metamorphism, recorded by 40Ar/39Ar and Rb/Sr cooling ages of 400–430 Ma.

In the Lochkovian Red Bay Group of the Raudfjorden Graben, two horizons of tuffites occur, interbedded with sandstones. New studies of eight zircons from these volcanic rocks have provided single-zircon lead-evaporation ages of c. 950 and c. 1350 Ma; one yielded 440 Ma. All these zircons are probably derived from the underlying basement rocks, the ages being significantly older than the Devonian host strata (c. 410 Ma).

The clasts in the Lilljeborgfjellet Conglomerate are generally angular to subrounded and derived locally from the underlying high-grade metamorphic complex. A subordinate (usually less than 1%, but up to about 10%) component of the clasts is a quartz porphyry that is not known in the exposed bedrock anywhere in northwestern Spitsbergen. The quartz porphyries are better rounded than the other clasts; however, the maximum diameter reaches 1.5 metres, indicating that transport distances are unlikely to have exceeded a few kilometres. Three quartz porphyry boulders have been dated by the single-zircon lead-evaporation method and shown to be of Palaeoproterozoic age, yielding ages of 1735±4, 1736±5 and 1739±5 Ma that have not previously been detected in the northwestern part of Svalbard's Caledonides.

The quartz porphyry clasts show no evidence of the widespread high-grade tectonothermal activity of Mesoproterozoic and early Palaeozoic age that influenced northwestern Spitsbergen. It is therefore concluded that the most probable source of these clasts lies to the east in the unexposed basement beneath the Old Red Sandstones of the Andrèeland–Dicksonland Graben. The Lilljeborgfjellet quartz porphyry clasts are closely similar in age to the granitic rocks of Ny Friesland. Whereas the latter were subject to Caledonian high amphibolite facies metamorphism, the quartz porphyry clasts have only been affected by a low greenschist facies overprint. Nevertheless, the similarity in age suggests an affinity to Ny Friesland and it is proposed here that the Breibogen–Bockfjorden Fault defines the most important boundary between Svalbard's Caledonian terranes.

Central Cape Breton Island in Nova Scotia, Canada, is host to ∼700–630 Ma felsic and associated mafic volcanic rocks that are relatively rare in other parts of the Avalon Composite Terrane, occurring elsewhere only in the Stirling Block of southern Cape Breton Island and in parts of eastern Newfoundland. The mafic rocks of central Cape Breton Island are typically intraplate tholeiitic basalts generated by melting of a garnet-bearing mantle source. They lack a continental trace element and εNd imprint although they were emplaced on continental crust; they resemble oceanic island basalts. Contemporaneous volcanism in the Stirling Block is calc-alkaline and formed in a volcanic arc setting. In the absence of evidence for an intervening trench complex or suture, it may be inferred that the central Cape Breton tholeiites formed in a back-arc setting relative to the Stirling Block. This rifting may represent the initial stages of separation of an Avalonian arc from western Gondwana. The arc rifted further between ∼630–610 Ma when the younger Antigonish-Cobequid back-arc basin formed. Subsequently, the extensional arc became convergent, telescoping the back-arc basin. Northwestward migration of calc-alkaline arc magmatism may be related to shallowing of the associated Benioff zone through time.

The western Koryak fold and thrust belt consists of a set of tectonostratigraphic terranes that contain units ranging from Lower Palaeozoic to Cenozoic. Three deformational events have been identified in the study area. The first event structures are folds, domes and shear zones with related high-pressure/low-temperature metamorphism. These structures are early Carboniferous and are only recognized in the metamorphic terranes. The second event structures are imbricate fans of thrusts and folds with southeast vergence, broken formation and serpentinite mélange. These are latest Jurassic to early Cretaceous (early Albian) and occur throughout the study area. During this event, thrusting was accompanied by dextral strike-slip faulting. The second event structures are overlapped by the Upper Albian sedimentary rocks with an angular unconformity at the base. During metamorphism associated with the first and second deformational events, some of the rocks were metamorphosed to blueschist grade and were affected by strain with axial ratios of up to 15[ratio ]1. The third deformational event is characterized by significant sinistral strike-slip displacement at higher crustal levels. This resulted in a new set of structures and rotation of pre-existing structures. The age of the sinistral strike-slip faults is interpreted to be late Cretaceous to Cenozoic. The kinematics of the second and third deformational events correspond to assumed proto-Pacific plate motions based on palaeomagnetic data.

Siberia contains several key reference sections for studies of biological and environmental evolution across the Proterozoic–Phanerozoic transition. The Platonovskaya Formation, exposed in the Turukhansk region of western Siberia, is an uppermost Proterozoic to Cambrian succession whose trace and body fossils place broad limits on the age of deposition, but do not permit detailed correlation with boundary successions elsewhere. In contrast, a striking negative carbon isotopic excursion in the lower part of the Platonovskaya Formation permits precise chemostratigraphic correlation with uppermost Yudomian successions in Siberia, and possibly worldwide. In addition to providing a tool for correlation, the isotopic excursion preserved in the Platonovskaya and contemporaneous successions documents a major biogeochemical event, likely involving the world ocean. The excursion coincides with the palaeontological breakpoint between Ediacaran- and Cambrian-style assemblages, suggesting a role for biogeochemical change in evolutionary events near the Proterozoic–Cambrian boundary.

New field, petrographic, geochemical and Rb–Sr/Sm–Nd isotope data are presented from early mafic dykes which intrude the Mesoproterozoic Cape Meredith Complex, West Falkland. The dykes, which have been previously regarded as being Ordovician in age, are seen in the field to cut a suite of lamprophyre sheets. New K–Ar data from biotite separates from two lamprophyres suggest a miminum age of emplacement of ∼520 Ma, confirming previous work. The Rb–Sr and Sm–Nd analyses of the pre-lamprophyre mafic dykes suggests that they were probably intruded during the late Neoproterozoic at ∼600 Ma. This new data has considerable implications for the previously published estimates of the longevity of intracontinental extension events in this part of Gondwana.

Recent discovery of wollastonite-bearing calc-silicate assemblages adjacent to gneiss–charnockite horizons in the supracrustal terrain of the Kerala Khondalite Belt, southern India, provides an opportunity to evaluate the carbonic fluid infiltration model proposed for charnockite formation. Petrological and fluid inclusion studies across these horizons in three representative localities are presented in this study. The calc-silicate assemblages define peak metamorphic conditions of ∼800°C at 5 kbar and define a low aCO2. Adjacent charnockite assemblages developed through dehydration involving the breakdown of garnet, biotite and quartz to produce orthopyroxene under low aH2O conditions. Retrograde reactions preserved in the calc-silicate rocks, such as scapolite–quartz symplectites, and the partial breakdown of wollastonite previously has been attributed to a near isothermal decompression during which infiltration of CO2-rich fluids occurred. Fluid inclusion studies indicate that the earliest generation of fluids preserved in the calc-silicate assemblages are aqueous (with salinity ∼8 wt% NaCl equivalent), consistent with mineral phase equilibria defining low aCO2. The estimation of NaCl content in brines coexisting with scapolite, based on the Cl content of the scapolite, indicates the presence of up to 20 wt % NaCl during the formation of scapolite consistent with the saline primary fluid inclusions. Primary carbonic inclusions occur within the retrogressed calcite+quartz assemblage after wollastonite, and are considered to represent the post-peak metamorphic carbonic fluid infiltration event, synchronous with the development of charnockites in the adjacent gneisses. These inclusions have identical characteristics to those in the charnockites. We envisage that the Kerala Khondalite Belt fluid regime was largely internally buffered during the prograde path, and that CO2 infiltration post-dated peak metamorphism. Influx of CO2 was mostly structurally controlled, and occurred along a near-isothermal uplift path. Graphite-bearing pegmatitic dykes with abundant CO2-rich inclusions in these localities attest to the transfer of carbonic fluids through magmatic conduits.

Integrated graptolite, conodont and chitinozoan biostratigraphical data are presented from the lower Silurian of the Ohesaare core, Saaremaa, Estonia. A precise correlation of graptolite and conodont biozonations for much of the Telychian and Sheinwoodian is presented; this differs significantly from correlations suggested previously, which were based on inadequate graptolite data. The boundary between the celloni and amorphognathoides conodont biozones lies close to that between the spiralis and lapworthi graptolite biozones. The boundary between the bicornis and procerus conodont superzones lies mid-way through the murchisoni graptolite Biozone. The top of the procerus conodont Superzone lies at or very close to the boundary between the murchisoni and firmus graptolite biozones. The boundary between the Lower and Upper ranuliformis conodont biozones lies close to the boundary between the firmus and riccartonensis graptolite biozones. The precise position of the Llandovery–Wenlock boundary cannot be identified in the core. The newly demonstrated biozonal correlations are significant for studies of Silurian chronostratigraphy (previous estimates of the duration of the epochs and ages of the Silurian were based on inaccurate correlations) and sea-level changes. Palaeogeographically, the setting of Ohesaare is considered to be relatively deep shelf. Perhaps surprisingly, there is a major unconformity in the Llandovery of the core: much, if not all, of the Aeronian is missing, as is the lower part of the Telychian.

Stable isotopic measurements have been made on both planktonic foraminifera and coccolithic matrix of Middle Cretaceous (Late Albian–Cenomanian) age from two Pacific low latitude sites. The degree of alteration of the foraminifera has been assessed through the application of chemical analyses, cathodoluminescence and Scanning Electron Microscopy (SEM). The rotaliporid foraminifera display an interspecies range of δ18O values from −2.29 to −3.01‰ at Deep Sea Drilling Project (DSDP) Site 463 and from −2.74 to −3.55‰ at DSDP Site 305. Hedbergellid foraminifera exhibit a δ18O interspecies variation of −2.52 to −3.02‰ at Site 305. Isotopic analysis of individual Hedbergella delrioensis and Rotalipora appenninica foraminifera from single samples shows H. delrioensis to have a surprisingly large spread of δ18O values (−2.492 to −3.097‰ from Site 463, −2.454 to −3.344‰ from Site 305), whilst δ13C values remain confined to a narrower range. Such a spread of oxygen values may be related to a number of factors, including subtle diagenetic alteration, a wide range of temperature-related depth habitats or growth related changes of primary skeletal calcite. The hedbergellids have consistently lighter oxygen and heavier carbon isotopic values than do the rotaliporid foraminifera and hence provide isotopically derived palaeotemperatures consistent with a thermally stratified ocean. At both sites the oxygen isotopic data are consistent with a gradual warming through Albian–Cenomanian time. However, the results suggest that Middle Cretaceous equatorial oceans were possibly only as warm as those of the present day (or slightly warmer), but did not reach the high temperatures claimed in older literature.

This paper consists of three parts. The first part presents a critical review of previous mechanical models on the relationship between joint spacing (s) and bed thickness (t) in sedimentary rocks. The second part describes a new mechanical model dealing with the effects of interbed slip on joint spacing. The third part presents the comparison between theoretical results of the model and joint data measured from Cambrian flysch sediments at Plage Victor in the Saint-Jean-Port-Joli area of the Quebec Appalachians. This study demonstrates two formation mechanisms of tensile joints: near-end fracturing takes place in the layers with smaller tensile fracture strength (C0<30 MPa) and smaller interbed shear strength (τ0<20 MPa) while mid-point fracturing occurs in those layers with larger C0 (>50 MPa) and τ0(>30 MPa) values. For the rocks with moderate tensile fracture strength (30<C0<50 MPa) and interbed shear strength (20<τ0<30 MPa), the near-end fracturing and mid-point fracturing take place preferentially in the thicker and thinner brittle layers, respectively. In the regime of near-end fracturing, the correlation between s and t is linear and independent of the thickness of the bounding non-jointing layers (d). In the regime of mid-point fracturing, however, the s–t relationship can be either linear or non-linear, depending on the variation of d value. The present study also suggests that the near-end fracturing is probably the prevailing process for the formation of tensile joints in bedded sedimentary rocks and that the coefficient of joint spacing (K) defined by the ratio of s to t is considered as an indicator of C0/(2τ0) for the jointed layer.

The widespread Triassic volcanic rocks of Greece, dismembered during the Hellenide orogeny, are used to interpret the nature of Triassic rifting. Four assemblages of volcanic rocks are distinguished on geochemical criteria: (1) a predominant subalkaline basalt–andesite–dacite series with a high proportion of pyroclastic rocks; (2) minor shoshonites; (3) alkali basalt and (4) MORB. The stratigraphic and palaeogeographic distribution of these rock types is synthesized. New Pb and Nd isotopic data are used to discriminate between hypotheses suggesting that either subduction or extension was responsible for the Triassic volcanism. In the subalkaline basalt assemblage, εNd is negative with depleted mantle model ages >1.5 Ga. Pb isotopic compositions are mostly close to the very distinctive compositional field of Cenozoic extensional rocks of the Aegean area, with very high 207Pb/204Pb for relatively low 206Pb/204Pb ratios. These isotopic data confirm interpretations based on trace elements that subalkaline basalts were predominantly derived from melt-depleted peridotite in the sub-continental lithospheric mantle as a result of extension. Small areas of enriched hydrous mantle partially melted to yield shoshonitic magmas. Nd and Pb isotopic compositions of the alkali basalts are quite different from those in other rock types and suggest a HIMU mantle source component derived from a small plume, which also influenced MORB compositions. Distribution of these various rock types is used to constrain palaeogeographic reconstruction of Triassic micro-continental blocks.

New δ13Corg analyses of two boundary sections between the late Permian Kapp Starostin Formation and the early Triassic Vardebukta Formation of western Spitsbergen confirm field evidence that their contact is a conformable one. Thus, contrary to previous reports, some Spitsbergen sections contain a complete record of the environmental and faunal changes during the crisis interval of the end Permian mass extinction. No environmental deterioration is recorded in the late Permian until near the end of the terminal Changxingian Stage, whereupon the abundant siliceous sponge fauna of the Kapp Starostin Formation disappears along with the deep-burrowing fauna responsible for the Zoophycus trace fossil. A low diversity dysaerobic trace fossil assemblage is briefly developed before a transition to finely laminated, pyritic facies immediately beneath the Permo-Triassic boundary. Analysis of the S/C ratios from the laminated strata suggests that free H2S was present in the water column (euxinic conditions) even in relatively nearshore settings subject to storm sandstone deposition. The mass extinction crisis in Spitsbergen is therefore coincident with the extensive development of oxygen-poor conditions in the water column and compares closely, both in timing and nature, with the crisis seen in lower latitude Tethyan settings. However, the subsequent aftermath and recovery in the Boreal sections of Spitsbergen was more rapid than in Tethys. Thus, a shoreface sandstone body within the Dienerian Stage contains an appreciable diversity of fauna (by the standards of the early Triassic), including bryozoans, calcareous algae and deep infaunal bivalves, that suggests the marine ecosystem recovery began earliest in higher palaeolatitudes.

One of the major metamorphic terranes of the Bohemian Massif, the Erzgebirge, is interpreted to record a subducted part of a Palaeozoic margin of Gondwana. A geochemical study on non-calcareous metasediments from the various metamorphic units from lower greenschist to granulite facies metamorphism supports a recently established thrust model. Geochemical discrimination and correlation from the metasediments of the Erzgebirge suggest repetition of an early Palaeozoic metasedimentary sequence in metamorphic thrust units. This new finding is in line with recent radiometric dating of intercalated metarhyolitic rocks, which yielded ages of around 480 Ma. It is furthermore supported by correlation with a low-grade standard section in Thuringia, which represents the transition from an orogenic belt to a passive margin setting, with highly mature sediments. Significant geochemical signatures have been identified in three different lithotypes, which reappear in at least three metamorphic units of the Erzgebirge. Geochemical correlation of these units was established using simple comparison of averages and with statistical techniques. The identification of significant geochemical signatures from different lithotypes in metamorphic suites has important implications for terrane analysis and reconstruction of ancient tectonic settings.

The repetition of lithologies and their distinct chemical compositions in progressively metamorphosed units is useful for examining element mobility during Barrovian metamorphism. Statistical comparison implies that Li is progressively depleted from the greenschist to amphibolite facies, whereas Ca exhibits some enrichment. All the other elements studied are considered to be immobile.