Anabaritids are a distinctive group of early skeletal fossils, with tri-radially symmetrical tubes. They occur in the earliest Cambrian deposits, and seem to be especially characteristic of the Tommotian strata and putative pre-Tommotian units (e.g. Manykay horizon). A detailed redescription of Anabarites rotundus Qian from near Kuanchuanpu, Shaanxi Province emphasizes the degree of morphological variation. New data on shell composition suggest a bilayered structure, the outer unit composed of fibres parallel to growth lines, overlying a distinctive inner lamella with a chevron ultrastructure. Tentative evidence for anabaritids possibly being operculate is presented, the putative operculum being a simple convex disc. More poorly preserved anabaritids from the Yangtze Gorges (Hubei Province) and Meishucun (Yunnan Province) are referred to A. trisulcatus. All these Chinese occurrences of anabaritids are consistent with a lower Tommotian age of the associated faunas, but reliable and precise correlations with other Precambrian-Cambrian boundary sections remain inconclusive.

We have investigated the reaction between crystalline amphibolite and hydrous granite melt in static experiments at 810 °C and 1.5 kbar. Boundary layer concentration gradients in quenched silicate glass for the major element oxides and the volatile components, water and carbon dioxide, were measured using electron probe analysis and Fourier Transform Infrared Spectroscopy, respectively. We found a measurable change in the concentration of all components adjacent to the amphibolite in experiments of 66 and 330 hours duration. After I hour there was no detectable change in the concentration of major element oxides in the granitic glass, but steep concentration profiles were determined for carbon dioxide and water. A bubble-free zone developed adjacent to the amphibolite in the 66 hour experiment, and this zone increased in width after 330 hours. Reaction is controlled by dissolution of amphibolite and by transport of dissolved material through the granite melt. The rate-controlling process is chemical diffusion in the melt phase. Results confirm that in the absence of convective heat transfer and/or mechanical disaggregation of mafic inclusions, assimilation of mafic rocks by granite melt is very slow, corresponding to on the order of 10 mm for SiO2 in 1000 years.

The Adola Fold and Thrust Belt of Ethiopia is a late- Precambrian, north–south trending belt of volcanic-sedimentary and ophiolite–like units overlying ‘basement rocks’ (gneisses and granitic gneisses). Detailed structural analysis and data from microfabrics have documented two thrusting events (D1, D5) and three folding phases (F2, F3, F4). All deformations have affected both the rocks of the Adola Fold and Thrust Belt and the ‘older basement’. The structural history is recorded as follows: (1) formation of ductile shear zones (Dl); (2) progressive development of recumbent folds (F2); (3) a superimposed north-south trending, tight to isoclinal, upright penetrative folding phase (F3), which imparted the dominant structure of the region; (4) F3 was later gently affected, at right angles to F3, by the last folding event (F4). Metamorphic conditions during this deformational cycle reached highest greenschist to medium amphibolite facies conditions. Close to major thrust contacts imbrication produced metamorphic discontinuities. Marked retrograde metamorphism in narrow horizons indicates a continuation of minor movements along the major thrusts, postdating the metamorphic peak. Evidence of a second thrusting/faulting episode (D5) is only recorded at the mesoscale. This latter event was accompanied by very low to low grade metamorphic conditions. In accepting the previously reported age of 1030±40 Ma for the volcanic–sedimentary unit of the Adola Fold and Thrust Belt, and the interpretation of these units as an immature island arc, a possibly early Pan-African oceanic accretion is postulated.

Strike, dip, and thickness were measured for 504 sheets (inclined sheets and dykes) in the 4–6 Ma old Hafnarfjall central volcano in southwest Iceland. The average dip of sheets is 65°, 80% are less than 1.2 m thick, and the thickness tends to decrease with decreasing dip. In 0.5 km long traverses perpendicular to the average strike of sheets, the percentage of sheets ranges from about 6 to 11.

Of 140 chemically analysed sheets most are quartz-tholeiites; a few are intermediate or acid. The sheets are chemically more evolved than the host rock and were generated by a shallow crustal magma chamber at a mature stage of the central volcano, whereas the host rock was generated earlier before the chamber was established. Trace element results suggest that the sheet magmas evolved by low-pressure fractional crystallization as well as by mixing of primitive magmas and crustal melts.

A model is proposed where most of the sheets are generated by a growing shallow magma chamber. As the chamber grows its shape changes, and so does the local stress field associated with it. Because the sheets follow the stress trajectories of the local stress field, the potential pathways of the sheets change with the growth of the chamber, which may explain the common occurrence of cross-cutting sheets. From the evolved chemistry of the sheets, as well as from the pattern of the stress trajectories, it is concluded that the bulk of the sheets were injected from the upper part of the shallow magma chamber.

The Bwlch y Cywion microgranite intrudes an Ordovician sedimentary sequence in northeast Snowdonia, North Wales. Its age is recorded by hornfelses within its metamorphic aureole which give an Ordovician age of 454 ± 20 Ma. The whole-rock Rb–Sr isotope systems of the intrusion, however, and of an associated ash-flow tuff and a rhyolite of the Llewelyn Volcanic Group (Caradoc), were reset during low-grade metamorphism and give Devonian ages of 392±11 Ma, 392±5 Ma, and 405 ± 6 Ma, respectively.

Basaltic hawaiite lavas are virtually absent in the Eocene Skye Main Lava Series, in contrast to relatively abundant basalts and hawaiites. Fractional crystallization from basalt to basaltic hawaiite involves extraction of a large proportion of plagioclase, and liquid densities thereby increase. From basaltic hawaiite to hawaiite titanomagnetite is a significant fractionating phase, and liquid densities decline. The coincidence between a gap in erupted compositions and a density maximum implies that liquid density exerted a strong control on ‘eruptibility’ of magmas; basaltic hawaiites were too dense to be erupted. Density maxima occur in basalt suites if plagioclase fractionates before Fe–Ti oxides, and may explain compositional gaps in erupted magmas. Compositional gaps are not the inevitable result of density maxima; the density of the rock column above, and the fluid dynamics within, the magma chamber where differentiation occurs are also critical factors.

If historical earthquake data are to be used in studies of regional seismotectonics, it is essential to include information on likely locational uncertainty. Contributory factors include both geophysical parameters and the social environment of the event. A reconstruction of how a recent earthquake might have appeared had it occurred 300 years ago is used to demonstrate the effects of documentary survival. Limits of uncertainty are estimated for two historical earthquakes (7 October 1690 and 28 November 1880) and shown to be large. It is recommended that such events should be shown on epicentral maps as areas rather than points.

Results are presented of thin-section, SEM, X-ray and microprobe investigations on problematical phosphatic accretions (‘ Bolopora ’) from the St Tudwal's area of North Wales. The data supplement investigations made by H. J. Hofmann on the genesis of the accretions, and indicate a microbial (bacteriogenic) origin. The bacteriogenic pebble coats are of dahllite. Microbial relics suggest water–sediment interface-linked formation of the coatings under subtidal, and oxic to anoxic, environmental conditions during a sedimentary hiatus.

Although the Mesozoic sediments of the Cleveland Basin (North Yorkshire) have generally not been strongly faulted, several approximately N–S trending faults have been identified along the coast. New seismic data from adjacent coastal waters has allowed the offshore extension to the fault system to be examined for the first time. The coastal faults from Peak (Ravenscar) to Red Cliff (Cayton Bay) are shown to form part of a linked system defining a narrow graben only some 5 km wide, the Peak Trough. Faulting has been complex, with decollement levels apparently developed in weak layers at various horizons in the Triassic and Permian strata: fault geometries and regional considerations suggest that extension has been dominant. Movement occurred intermittently from Triassic to latest Cretaceous or early Tertiary times.

An acritarch assemblage from beds beneath the Ogof Hên Formation near Carmarthen (the ‘Login beds’) resembles assemblages from the Watch Hill Grits and underlying sediments of latest Tremadoc or earliest Arenig age in the Skiddaw Group of the English Lake District. We conclude that the Login acritarch assemblage is younger than at least some part of the late Tremadoc Angelina sedgwickii Biozone, but older than the early Arenig Didymograptus deflexus Biozone. Its occurrence may therefore coincide with that of the graptolite Tetragraptus approximatus. This is of particular significance in any attempt to define an internationally correlatable base for the Arenig Series in Britain.

A previously unrecorded quartz-dolerite dyke some 40 km long and up to 15 m wide intrudes Lower Palaeozoic rocks in southwest Wales between Mathry and Llanboidy. It is of discontinuous subcrop but has an overall WNW trend. This crosscuts the Caledonian strike of the country rocks but is parallel with the line of the Hercynian front to the south. The dyke is probably related to the suite of Dinantian basalts and dolerites near the Bristol Channel, to the east of the South Wales Coalfield.

A more precise age for the synkinematic emplacement of the large S-type Leinster batholith (southeast Ireland) has been obtained by U–Pb dating of zircon and monazite separates from the two main granite varieties of the Northern Pluton. A concordant age of 405±2 Ma derived from monazite is considered to reflect both the age of intrusion and the main regional deformation event (D1) in southeast Ireland. The monazite age is in good agreement with an earlier, less precise whole-rock Rb–Sr isochron determination. The zircon data indicate recent Pb loss effects and the existence of a small component of inherited radiogenic Pb derived from a crustal precursor during the generation of the batholith.