The onshore record of Cenozoic glaciation in the Antarctic Peninsula region is limited to a number of isolated localities on Alexander Island, the South Shetland Islands and in the James Ross Island area. In the James Ross Island area, Late Cretaceous sedimentary rocks are unconformably overlain by a unit of diamictites and tuffs, which occur at the base of the James Ross Island Volcanic Group. These rocks are here defined as the Hobbs Glacier Formation, and on the basis of palynological studies are assigned to a Miocene (?late Miocene) age. The diamictites are interpreted as representing glaciomarine sedimentation close to the grounding line of either a floating ice shelf or a grounded tidewater glacier in a marine basin. Provenance studies indicate that the glacier was flowing from the Antarctic Peninsula towards the southeast. Volcanic tuffs conformably overlie the diamictites and are interpreted as representing deposition in a periglacial delta front setting in either a marine or non-marine basin, away from direct glacial influence. The Hobbs Glacier Formation and overlying James Ross Island Volcanic Group help to enhance our understanding of the Neogene glacial chronology of West Antarctica.

Four hundred and eighty-four whole rock dolomite samples from early Proterozoic strata have been analysed for carbon isotopes. The carbon isotope composition is strata-specific, but abrupt variations have been observed between the Jianancun and the Daguandong as well as the Daguandong and Huaiyincun formations. The carbon isotope composition of dolomites is believed to reflect δ13C of oceanic total dissolved inorganic carbon during the early Proterozoic and reflect changing rates of organic carbon burial.

Spinel lherzolite and harzburgite xenoliths hosted in an alkali basalt dyke near Inver, Donegal, Ireland show abundant evidence of interaction between xenolith minerals and the host melt. Of particular interest are primary Cr-diopside and spinel with sieve-textured coronas. Coronas on primary Cr-diopside are up to 3 mm wide and are associated with veinlets of devitrified glass. The coronas comprise secondary Cr-diopside with vermicular, interstitial alkali feldspar and chlorite grains up to 100 µm in size. The inclusion-free Cr-diopside cores are Al- and Na-rich whereas the coronas are Al- and Na-depleted and Ti-enriched. Sieve-textured spinels have similar texture to the clinopyroxene grains and are also associated with veinlets of infiltrated glass. However, the interstitial inclusions in the sieve-textured region are chlorite and nepheline. Inclusion-free spinel is part of a chromite–spinel solid solution and is Ti-poor. Spinel in the coronas has a greater chromite and ulvospinel component and falls close to a mixing line with spinel in the host alkali basalt. In addition to the sieve-textured grains, primary olivine in contact with infiltrated glass has Fe-rich rims, and orthopyroxene has broken down to form rims of olivine, clinopyroxene and a K-rich phase similar in composition to alkali-feldspar. Comparison of the compositions of the inclusion-free cores and sieve-textured rims shows that the rims have chemical signatures consistent with partial melting, that is, Al and Na depletion for clinopyroxene and Cr-enrichment for spinel. The textures of the coronas, particularly those around spinel and the reaction margins on orthopyroxene are identical to those produced during dissolution experiments.

We suggest that silicate liquid from the host magma infiltrated the xenoliths during their ascent and since it was not in equilibrium with the xenolith minerals caused reaction. The occurrence of K-bearing interstitial minerals in the sieve-textured grains and reacted orthopyroxenes indicate that the coronas did not form by simple melting since none of the minerals that underwent breakdown are K-bearing. We suggest that the sieve-textured grains formed initially by partial melting and reaction associated with decompression and infiltration of liquid from the host magma. The melts included in the reacted phases were enriched in K by diffusion from the Si-poor infiltrated melt into the more Si-rich melt inclusions in the coronas.

The Flagstone Bench Formation ranges in age from earliest Triassic to Norian (Late Triassic) and is exposed in the Beaver Lake area of the northern Prince Charles Mountains. This sandstone-dominated formation rests conformably on the Bainmedart Coal Measures and represents the upper part of the Permian–Triassic Amery Group. It is divisible into three members: the Ritchie, Jetty and McKelvey members (in ascending order). Nine sedimentary facies assignable to three facies associations (major channel, crevasse/fan and flood-basin deposits) are recognized within the formation. Ritchie Member sedimentation took place during a transition from consistently hygric to seasonally dry conditions and the unit comprises sandstone-dominated, sheet-like channel deposits interspersed with few, thin, mottled, haematite-rich flood-basin siltstones. Deposition took place under fluctuating discharge conditions chiefly within the channel tracts of axially (northwesterly/northeasterly) flowing, low-sinuosity braided rivers. The Jetty Member shows a gross upward-fining profile dominated in the lower part by poorly sorted pebbly sandstones and in the upper part by ferruginous mudcracked siltstones, mottled palaeosols, calcrete and thin massive sandstone sheets. This unit reflects deposition of easterly directed alluvial fans and extensive flood-basin silt under a semi-arid climatic regime. The Upper Triassic sandstone-dominated McKelvey Member shows a return to axial drainage along the Lambert Graben with sedimentation occurring primarily within low-sinuosity braided channel tracts under wetter climatic conditions.

Preliminary multidisciplinary investigations were carried out in a 105 m thick late Pleistocene terrestrial sequence in the Lamayuru basin, Ladakh Himalaya. Palaeomagnetic studies reveal a reversal of polarity at about 35 ka BP. This is the only report of this reversal event in the late Pleistocene terrestrial sequences of the Indian subcontinent. The valley fill sequence is the product of an interplay of lacustrine with fluvio-deltaic to colluvial processes. Four fossiliferous horizons have yielded ostracods and gastropods indicating shallow freshwater conditions. The Lamayuru lake was created by tectonically induced instability about 35–40 ka BP, an event that may be represented widely throughout the Himalaya.

Remigolepis walkeri sp. nov. from the Mandagery Sandstone (late Devonian) near Canowindra, New South Wales (NSW), is the second species of Remigolepis to be described from Australia, the first being from near Grenfell, NSW. Remigolepis walkeri possesses unusual paired suborbital plates with a large oval structure at the anteromesial edge of the plate, representing an attachment for the autopalatine portion of the palatoquadrate. Among asterolepidoids, this morphology is most similar to Pterichthyodes. Suborbitals of Remigolepis from East Greenland are said to possess a transverse ridge on the internal surface, similar to the bothriolepids Bothriolepis and Nawagiaspis. However, some specimens from East Greenland may show a morphology more similar to Remigolepis walkeri. The internal morphology of the suborbital plates is constant in the population of Remigolepis from the Canowindra locality, suggesting the presence of a single species despite the presence of more than 1000 individuals in this fauna. The morphology of the caudal fin of Remigolepis walkeri is similar to Remigolepis sp. from near Eden, NSW, and Asterolepis ornata, but differs from Remigolepis sp. from China. The morphology of the pectoral and caudal fins of R. walkeri indicate a bottom-dwelling lifestyle, whereas Bothriolepis from the same fauna may have been able to generate sufficient lift from the pectoral fins to enter the water column on a regular basis.

A 5 m long permineralized trunk from the Upper Kellwasser member of southern Morocco represents the first record of a large trunk of an identifiable species of Callixylon, C. erianum, from Gondwana. This occurrence constitutes the most reliable evidence based on plant megafossils for a floral connection between Laurussia and Gondwana in late Devonian times and for a proximity of these continents in Famennian times. The potential of this trunk for studies of the architecture and growth patterns of the earliest trees with a gymnospermous type of arborescent habit is discussed.

The Upper Carboniferous Kent Coalfield lies concealed beneath various Mesozoic formations, its southern areas lying about 20 km north of the commonly accepted position of the main Variscan Deformation Front. However, despite intense intra-coal deformation, the existing literature is ambivalent about compressional Variscan features in Kent, the general view being that coal deformation is largely the product of the depositional environment. The main deformation is interpreted here as the result of Variscan compression, the structural style being imposed by the sandstone-dominated lithology. This conclusion is necessitated by the regularity of deformational structures revealed by mine workings, and supported by coal sequence irregularities suggestive of thrusting, especially in the lower Westphalian strata, all of which is paralleled in parts of the South Wales Coalfield. The Kent data indicate that, as in South Wales, a zone of thrusting many tens of kilometres wide lies in advance of the main deformation front. Structural trends are consistent with an overall swing in the front from east–west across much of central-southern England, to more northwest–southeast across northeastern France. This swing may represent a transpressional transfer zone, within which stress deflection and block rotation produced thrust vergence oblique to the overall direction of maximum compression.

Radiolarian and calpionellid dating establish a significant sedimentary hiatus within a reduced pelagic sequence of the central part of the Ionian zone (western Greece). Bottom-current activity probably decreased during the Oxfordian, allowing encroachment of radiolarian-rich sediments onto a Jurassic high (at Kato Kouklessi), and then increased again during Kimmeridgian–Tithonian time. The onset of Vigla Limestone sedimentation is diachronous within the Ionian zone, spanning late Tithonian to various levels of early Berriasian time. This diachroneity can be explained by dispersal of fine-grained sediment by current activity.

M. B. Edwards comments: Of all the geological occurrences he had seen, this was the most remarkable and revealing. For him it was a spiritual place, and the journey was a pilgrimage; he used this word himself. This was now the seventh time he was making this journey. Thus wrote Rosendahl (1945, p. 22; my translation from Norwegian) about the geologist Sederholm. Decades later, after dozens more publications of innumerable additional outcrops, the ‘Reusch's Moraine’ exposure of upper Proterozoic glacial tillite resting on a glacially striated pavement (Reusch, 1891) remains unique and significant.

Writing about this exposure, Jensen & Wulff-Pedersen (1996) concluded ‘There is no evidence for glacial origin or contribution for the diamictite material’ (p. 137) in the Bigganjargga tillite. This discussion explains why I disagree with their conclusions.