The nucleation of barite has been studied in a system involving the doublediffusion of Ba2+ and SO42- in an Na-rich aqueoussolution, through a porous medium. The evolution of the concentration profiles in the medium, which is a column of porous silica gel, has been determined as a function of time by direct chemical analysis of the diffusion-controlled mass transfer. By measuring the pH evolution, a Debye–Huckel treatment of ionic complexing has enabled the supersaturation evolution to be determined. The location of barite precipitation in the column is controlled both by the need to exceed a threshold supersaturation, as well as achieve an ‘equality range’ in which [Ba2+]/[SO42−] is close to unity. The value of the threshold supersaturation is a kinetic parameter and depends on the rate at which supersaturation increases. The experimental system described here has wide application to the study of crystallization phenomena in rocks. Experiments on the effect of additives designed to inhibit nucleation of barite in North Sea oil wells are used to quantify the resultant increase in supersaturation threshold.

Fourteen early Ordovician (Tremadoc-Llandeilo) graptolite sequences from around the world are precisely (infrazonally) correlated, based on the stratigraphic ranges of 130 species and species groups. The composite standard sequence (CSS) of graptolites has been determined from the six best regional sequences by a nonparametric graphic correlation. Two data sets were selected: one comprised first appearance events of 103 taxa, the other, first and last appearance events of 45 taxa. The results of the two runs accord well and reveal respectively 66 and 73 successive bioevents in early Ordovician time. Event spacing averages 0.7–0.8 Ma and enables fine subdivision, correlation and homotaxial testing for diachroneity. The strong correlation between each of the six regional sequences and the CSS indicates the high level of accordance among graptolite successions around the world.

The evidence is reviewed for the timing of collision between the microcontinent of Eastern Avalonia (southern Britain and adjacent areas) and the Laurentian continent. Recent palaeomagnetic results placing Eastern Avalonia in a high (50°) southern latitude in mid Ordovician time are now consistent with faunal evidence for the first time. The resulting apparent polar wander path is evaluated and suggests that Eastern Avalonia detached itself from a southern peri-Gondwanan latitude in the early Ordovician, moved northwards, and approached Laurentia by the late Ordovician. Its western corner probably impinged on Laurentia in the early Silurian and it docked against the Laurentian margin during Silurian and early Devonian time with a component of anticlockwise rotation.

This kinematic history is supported by a compilation of sediment dispersal patterns on Eastern Avalonia. A low-volume Ordovician and earliest Silurian supply from within the microcontinent was overwhelmed in late Llandovery time by a large volume of southwest-derived turbidites, probably from the uplifting impact zone to the west. This source was later augmented by a high-volume clastic supply to the north margin of the microcontinent. Eastward migration of this source through Wenlock and Ludlow time reflects the progressive anticlockwise docking of Eastern Avalonia against the Laurentian margin. The earliest sign of a large-volume supply from Baltica is in the late Wenlock, arguing against any earlier hard collision.

Ultramafic xenoliths from Madeira island are divided into dunite/websterite/wehrlite/clinopyroxenite (DWWC) and harzburgite/lherzolite suites; the harzburgite/lherzolite xenoliths show abundant deformational features and are more refractory (Fo = 90–91) than the DWWC suite (Fo = 77–88).

DWWC xenoliths are spinel-bearing olivine ± orthophyroxene cumulates with intercumulus clinopyroxene and rare plagioclase, amphibole and phlogopite. Mineral chemistry and geothermobarometric data indicate that DWWC xenoliths crystallized at 1150–1300 °C from Madeiran alkalic basalts and accumulated in magma reservoir(s) located 36–45 km beneath the island.

The harzburgite/lherzolite xenoliths are composed of olivine + orthopyroxene + spinel ± clinopyroxene ± (rare) phlogopite and display alkali feldspar or clinopyroxenite veins and crystal aggregates. The complex thermal evolution recorded by these xenoliths and the close similarity of clinopyroxene REE contents and calculated fO2 values in both harzburgites and DWWC cumulates are attributed to recent infiltration of the harzburgites by melts trapped or crystallized within the mantle; these features, and the refractory bulk chemistry of the harzburgite/lherzolite suite, support the interpretation that these xenoliths represent depleted oceanic lithosphere variously modified by magmatism associated with the genesis of Madeira island. The association of these upper mantle xenoliths with cumulates crystallized from Madeiran magmas (DWWC) suggests that the harzburgite/lherzolite suite originated in the uppermost mantle above magma storage zone(s), probably near the boundary between the mantle and the overlying oceanic crust.

The Chuping Limestone of northwest Peninsular Malaysia, until recently considered entirely of Permian age, has yielded late Triassic (early Norian) conodonts. TheLimestone thus spans the time interval late Early Permian–Late Triassic and is in part equivalent to the Kodiang Limestone (Late Permian–Late Triassic) in Kedah andsimilar limestone sequences in south Thailand and north Sumatra. Early Late Triassic (Carnian) conodonts are also reported from pelagic limestones associated with bedded chertsof the Chert Member of the Semanggol Formation in Kedah. The Chert Member, previously considered of Middle Triassic age, is re-interpreted to represent Early, Middle and early Late Triassic deposition. The Triassic sedimentary rocks of the Malay Peninsula represent three distinct sedimentary regions: a stable shallow marine carbonate complex (ChupingLimestone, Kodiang Limestone), which forms part of an elongate carbonate platform on theSibumasu block; a deep water pelagic/turbidite basinal sequence (Semanggol Formation) which accumulated in either a foredeep basin or an intracratonic pull-apart basin related to strike-slip faulting; and a volcanic-sourced volcaniclastic basinal sequence on the East Malaya block (Semantan Formation and equivalents) which accumulated in either a forearc/intra-arc setting, or in a post-orogenic rift basin.

Granitic rocks from all the exposed components of the Lake District batholith have been analysed for ammonium to look for evidence of an ammonium-rich sedimentaryprotolith. The results indicate that the ammonium content of Lake District granites is related to the type and degree of alteration experienced by the rocks. Fresh, little-altered granites from all parts of the batholith have low ammonium contents (< 30 ppm), whereas highly altered rocks contain up to 250 ppm ammonium. The most ammonium-rich rocks are highly altered specimens of Skiddaw Granite containing appreciable secondary muscovite, sericite and clay. The degree of ammonium enrichment is related to the style of alteration and host-rock lithology. Evidence for the source of ammonium is provided by strong depletion (< × 0.3) of ammonium in silty mudstones within the cordierite andgarnet hornfels zones of the Skiddaw Granite aureole. In view of the low ammonium content of some sediments and the losses known to accompany high-grade metamorphism and associated dehydration, the low primary ammonium contents of the granites are not considered to be evidence either for or against a sedimentary protolith.

A synonymized nomenclature for calcified sponges (archaeocyaths, stromatoporoids, chaetetids, sphinctozoans and pharetronids) is here presented to enable comparison of suggested homologous structures.

A magnetic survey was carried out along a NW–SE trending monchiquiteintrusion near Llanllowell, Gwent. The intrusion was previously thought to be a dyke, 23m long and 5 m wide, and that it may have been an extension of a second monchiquite intrusion at Great House to the southeast. However, the survey revealed the intrusion to beat least 34 m long, up to 20 m wide, and independent of the Great House intrusion. Evidence suggests that both of these monchiquite intrusions in central Gwent may have been feeder pipes to post-Lower Carboniferous volcanoes.

The Boom Clay Formation of northeast Belgium consists of a sequence of alternating, laterally persistent beds of silt and clay. Power-spectral analysis of grain-size variations indicates regular sedimentary cycles with wavelengths of about 100 and 46cm that had periods of 124 and 57 thousand years (ka) or less. The lateral continuity ofthe beds, combined with the regularity and estimated periods of the cycles, suggests an indirect link to the 100 and 41 ka orbital (Milankovitch) cycles. The variations in meangrain size were caused by changing bottom-water turbulence as controlled by storms and/or glacio-eustatic water depth fluctuations. This study confirms that certain marinesiliciclastic sequences can provide proxy-records of palaeoclimatic variance.