Evidence is presented that many of the major strike faults in the British and Irish Caledonides were active as sinistral strike-slip zones in the end-Silurian to pre-mid-Devonian period. Some, such as the Highland Boundary Fault, moved in this way at an earlier stage in the Ordovician. These data allow the Caledonian rocks lying between the Laurentian miogeocline (whose basement is represented by the Lewisian, Moine and possibly the Dalradian) and the Gondwanaland miogeocline (Midland Platform and Welsh Basin) to be re-analysed as a group of disorganized terranes which originated to the southwest in North America and southwest Europe/Africa prior to the Silurian. The Highland Border Terrane and Northern Belt Terrane are interpreted as duplicated pieces of a mid-Ordovician sequence which was a back are to northwest subduction. The Midland Valley Terrane is interpreted as a slice of Laurentian foreland onto which ophiolites were obducted in the lower Ordovician but which became the basement of a continental margin arc to northwest subduction in the mid-Ordovician. The Cockburnland Terrane is inferred to be part of the same arc repeated and then broken up and dispersed by continuing strike slip. The Connemara Terrane is regarded as an allochthonous piece of the Dalradian miogeocline and the South Mayo Terrane as a remnant of an early Ordovician arc and fore arc which in mid-Ordovician times became a back arc/marginal basin to northwest subduction. The Lake District-Wexford Terrane is part of an arc to southeast subduction under Gondwanaland whose activity climaxed in the mid-Ordovician. The Central Terrane is interpreted as a Silurian overstep assemblage which blankets the junction between Laurentian- and Gondwanaland-derived oceanic terranes, and therefore Iapetus is regarded as an Ordovician ocean which closed prior to the Silurian. The model suggests that at the end of the Silurian, a clockwise-rotating Gondwanaland, having carried Laurentia into collision with Baltica, broke free and created a major sinistral strike-slip zone which disrupted the Ordovician palaeogeography in the British Isles/North American sector of Iapetus.

The Pitts Head Tuff Formation, of Ordovician (Caradoc) age, was emplaced as a thick (c. 700 m) intracaldera sequence and two outflow units comprising welded acidic ash-flow tuff. The Pitts Head pyroclastic flows were erupted subaerially but the lower and most extensive of the outflows crossed a shoreface, and continued for several kilometres offshore. The flow entered the sea without disruption and, following deflation and tuff emplacement, displaced the shoreface several kilometres to the east and northeast. Post-eruption subsidence in the northeast resulted here in the rapid establishment of environments deeper than had previously existed.

The lower outflow tuff is parataxitically to eutaxitically welded in both the subaerial and marine environments. The extremely regular plane-parallel welding foliation of the subaerial tuffs, however, contrasts with the locally highly deformed foliation of the tuff deposited beyond the shoreface. The deformed foliation, associated with irregular zones of intense siliceous nodule development, is ascribed to the upward streaming of water vapour generated at the tuff/sediment boundary. Elsewhere rheomorphism within the tuff was caused by instability resulting from emplacement on slopes related to faulting. Continued movement initiated extensive brecciation, detachment, and local gravity sliding of large rafts of tuff.

In northern Greenland in early Palaeozoic time a turbidite trough (the eastward extension of the Hazen trough of Arctic Canada) was flanked to the south by a carbonate platform. The trough was deformed during the mid-Palaeozoic Ellesmerian orogeny to form the E–W trending North Greenland fold belt. This fold belt was deformed further by Eurekan (Tertiary) structures, important among which is a major fault complex, the Harder Fjord fault zone (HFFZ). The suggestion has been made that this fault zone controlled early Cambrian sedimentation, even though the fault trace does not coincide with the trough–platform facies transition in sediments of that age; this has led to some controversy.

We report new information from a mapping programme by the Geological Survey of Greenland which has established the thin-skinned nature of Ellesmerian deformation at the trough-platform transition and implies that much of the fold belt is underlain by a shallow detachment. This in turn implies that the HFFZ exists in the hanging-wall of the detachment while the early Cambrian trough-platform transition is located autochthonously in the foot-wall. We adduce evidence to show that the latter was probably controlled by syndepositional faulting with actively eroding fault scarps and suggest that these basement structures were reactivated in a dextral strike-slip mode in early Tertiary time to form the HFFZ as now observed.

The relationship between disposition of metamorphic isograds and early strain history is examined for the Sulitjelma Fold Nappe of the central Scandes. Evidence for stratigraphic inversion and the widespread inversion of the garnet isograd is reviewed, and new evidence for the nature of early Scandian deformation is presented. A two-stage model is presented which explains the inversions by (i) an initial phase of horizontal compression resulting in crustal thickening by vertical stretching during closure of a marginal basin, followed by (ii) continued horizontal compression and crustal thickening by eastward-directed nappe emplacement. The emplacement of the Sulitjelma Fold Nappe during the latter phase of progressive simple-shear-dominated deformation was accompanied by the development of sheath fold geometries and the overturning of isotherms to produce the widespread inversion of the garnet isograd.

Evaporites within Upper Triassic to Lower Cretaceous sequences in the Atacama region of northern Chile are interpreted as the deposits of continental and coastal saline lakes. Halite casts and finely laminated calcareous evaporites, intercalated with alluvial and fluvial sediments, are probably playa lake deposits. These continental evaporites have been recognized in Upper Triassic alluvial sediments (Cifuncho Formation), in Upper Triassic–Lower Jurassic braided river deposits (basal unit of the Pan de Azúcar Formation), and within a sequence of Lower Cretaceous aeolian, alluvial and playa lake mudflat sediments (Quebrada Monardes Formation). Lower Cretaceous evaporites between marine limestones (Lautaro Formation) and continental redbeds (Quebrada Monardes Formation) were probably deposited in coastal saline lagoons, produced during a regionally extensive marine regression.

These sequences, and other similar successions in northern Chile, provide a record of almost continuous evaporite deposition, and hence of arid to semi-arid conditions, since Upper Triassic times. These conditions were primarily the result of a constant latitudinal position within the subtropical zone. A contributary factor was the geographical position of the area, initially on the west coast of Gondwanaland and subsequently on the coast of South America, with cold, northward-flowing ocean currents and offshore winds.