Ten sections of Neogene molasse-type sediments were sampled in the Guide Basin of northeastern Tibet for magnetostratigraphy [X.M. Fang, M.D. Yan, R. Van der Voo, D.R., Rea, C. Song, J.M. Parés, J. Gao, J. Nie, S. Dai, Late Cenozoic deformation and uplift of the NE Tibetan plateau: evidence from high resolution magnetostratigraphy of the Guide Basin, Qinghai Province, China, Geol. Soc. America Bull. 107 (2005) 1208–1225 [1]], but they also yield seven well-dated formation-mean directions that reveal changing declinations as rotations occurred in response to crustal deformation north of the India–Asia collision zone. Three formations are of early Miocene and Oligocene age, as indicated by fossils and magnetic reversal records, whereas four younger formations yield late Miocene and Pliocene ages. The dual-polarity magnetizations are typically antipodal, but reveal inclinations that are too shallow, most likely because of post-depositional inclination flattening. The late Miocene and younger directions show formation-mean declinations between 354° and 7°, whereas three early Miocene and late Oligocene mean declinations range from 31° to 44°. This indicates that a clockwise rotation of 25.1±4.6° took place during the middle part of the Miocene (best estimate 11–17 Ma). No rotations appear to have occurred, during that time, in the Xining, Lanzhou, Linxia and Jingning basins (Longzhong Basin) to the northeast and east of the Guide Basin; however, a rotation of similar magnitude has been documented by Dupont-Nivet and colleagues for pre-Miocene (>29 Ma) time in these areas. Collectively, these results show that the basins in NE Tibet have had independently evolving structural histories.

Paleomagnetic data from northern Appalachian terranes identify several arcs within the Iapetus ocean in the Early to Middle Ordovician, including a peri-laurentian are at similar to 10 degrees-20 degrees S, a peri-Avalonian are at similar to 50 degrees-60 degrees S, and an intra-oceanic are (called the Exploits are) at similar to 30 degrees S. The peri-Avalonian and Exploits arcs are characterized by Arenigian to Llanvirnian Celtic fauna that are distinct from similarly aged Toquima-Table Head fauna of the Laurentian margin, and peri-Laurentian are. The Precordillera terrane of Argentina is also characterized by an increasing proportion of Celtic fauna from Arenig to Llanvirn time, which implies (1) that it was in reproductive communication with the peri-Avalonian and Exploits arcs, and (2) that it must have been separate from Laurentia and the peri-laurentian are well before it collided with Gondwana. Collectively, the paleomagnetic and faunal data require an open Ordovician ocean adjacent to the Appalachian margin and argue against a Taconic-Famatinian collision between North and South America.

Scanning and transmission electron microscopy (STEM) of ocean-floor pillow basalts with ages between 0 and 70 Ma reveals progressive alteration of submicrometer titanomagnetite to phases such as goethite and clays. In contrast, larger titanomagnetite grains (> 1 μm) oxidize to titanomaghemite without apparent change in crystal morphology. Rock magnetic experiments are consistent with a selective removal of the submicrometer grains as the basalts age, and show good correlations between the ranges of grain sizes observed by STEM and those indicated by hysteresis properties. Remanence contributions from the larger (pseudo-single domain and multi-domain) titanium-iron oxides are inferred to decrease only slightly as the ocean floor becomes older, whereas the overall remanence decays with age as the substantial contribution from stable, single-domain titanomagnetite grains diminishes greatly due to their alteration to other phases. Although more work on many more samples is required to verify our conclusions, the current data imply that this alteration is one of the reasons that amplitudes of marine magnetic anomalies diminish with age over time scales of tens of millions of years.

A detailed paleomagnetic study in a single, first‐order, refolded syncline in the western hinge zone of the Variscan foreland fold and thrust belt of northern Spain was carried out to constrain the tectonic deformation of the Cantabrian Arc. Results show a complex pattern of structural deformation and acquisition of at least two secondary syntectonic magnetizations. On the basis of local and regional fold tests we document an Early Permian (B) component that postdates earlier (F1) Westphalian and Stephanian thrusting and folding and a Late Carboniferous to earliest Permian (C) component that is synfolding, acquired during Fl. The B component only resides in the carbonates and is carried by magnetite. The C component is carried by magnetite in the gray carbonates and by hematite in the red beds. Both B and C components reveal significant vertical‐axis rotations that indicate that an originally cylindrical Lagos del Valle Syncline was refolded (F2) about a vertical axis after acquisition of the B component magnetization. The pattern of rotations suggests that the entire Cantabrian Arc may have originated as a much more linear fold and thrust belt, which was later modified by significant oroclinal bending with relative rotations totaling at least 120°. The temporal distribution of remagnetization and deformation also suggests a protracted Variscan orogeny in the area, lasting from at least the Late Carboniferous into the Permian.

The origins of multiple magnetizations of the Archean Stillwater Complex have been investigated through scanning electron microscopy and scanning transmission electron microscopy observations of mineralogical relations, using representative samples from nine sites in mafic Banded series rocks. On the basis of directional grouping and demagnetization behavior, three magnetizations (here labeled A, B, and C) have been recognized. The natural remanent magnetizaton (NRM) is typically dominated by only one of these magnetizations and multicomponent behavior in individual specimens is rare. The A remanence resides in magnetic grains of high median destructive fields and high, discrete laboratory unblocking temperatures and is inferred to be a primary thermoremanent magnetization, of circa 2.71 Ga age. The B and C magnetizations, of lower median destructive fields and more distributed unblocking temperatures, are inferred to be secondary and related to alteration, including serpentinization, involving limited, moderate to low-temperature fluid interaction, perhaps in response to thermotectonic events (e.g., mafic dike emplacement). Samples with NRM dominated by the A magnetization contain titanium-free magnetite needles (width < 1 mu m and maximum dimension between about 2 and 50 mu m) with preferred orientations in cumulus plagioclase crystals; these oxides are interpreted to have formed during initial crystallization of cumulus phases. In samples containing only B or C magnetizations, such needles are absent in cumulus plagioclase, whereas secondary Fe-bearing silicates, particularly zoisite and chlorite, are conspicuous. The samples dominated by the B or C magnetizations contain anhedral low-chromian magnetite (FeCr2-xFexO4, 1.5 < x < 2.0) along fractures in both cumulus plagioclase and pyroxene and in zones of more pervasive alteration of pyroxenes. The paragenesis of these oxides is consistent with an origin related to hydrothermal alteration. Hysteresis data show that plagioclase crystals containing magnetite needles have higher remanent coercivities (H-cr = 60 to 100 mT), whereas plagioclase crystals with anhedral chromian magnetite have remanent coercivities less than 60 mT, consistent with their alternating field response. Because oxygen isotope data show no indication of appreciable alteration of host plagioclase, we infer that the absence of magnetite needles and their associated A magnetization are caused by selective alteration of the magnetite needles to nonmagnetic phases in the remagnetized gabbros. In this case, rock magnetic and paleomagnetic data appear to be higher resolution indicators of fluid/rock interaction than the stable isotope data. Anhedral chromian magnetite precipitated during hydrothermal alteration and therefore carries secondary magnetizations.

High resolution electron microscope studies have been carried out on 'zero-age' (New Flow) basalts from the Juan de Fuca Ridge and on young (< 20 ka) basalts from the axis of the East Pacific Rise at 12°N. Such data lead to characterization of the magnetic minerals, especially those of smaller grain size, which have been hypothesized by Kent and Gee to have undergone grain size-dependent alteration. In addition to larger titanomagnetite grains, abundant submicrometer titanomagnetite has been observed in globules within a glassy matrix. These grains, likely to be single-domain (SD) or superparamagnetic, are associated with apatite, uncommon pyrrhotite and residual glass. The submicrometer titanomagnetite grains have a wide compositional range (0 < x < 0.8), where x is the fraction of ulvöspinel component, whereas the larger, multi-domain (MD)-sized titanomagnetite grains have a narrow composition range of approximately x = 0.6. This variability in Ti content provides a ready explanation for the thermal rock magnetic properties observed by Kent and Gee and eliminates the need to invoke extremely rapid (< 20 ka) alteration of these young basalts.

The formation of the supercontinent Gondwana heralded the beginning of the Phanerozoic following a complex series of collisional events after the break-up of earlier supercontinental assemblages. Paleomagnetic data are used to help distinguish between these events and it appears that there are three critical periods of mountain building during Gondwana assembly. The first major orogenic event took place between 800 and 650 Ma and has been termed the East Africa Orogeny. This tectonic episode formed the Mozambique Belt and likely resulted from the collision of India, Madagascar and Sri Lanka with East Africa. The second and third orogenic periods during Gondwana assembly partially overlap in time. The Brasiliano orogeny (600–530 Ma) resulted in the amalgamation of the South American nuclei and Africa. The Kuunga Orogeny was proposed, in part, because of the recent collection of geochronologic data indicating a 550 Ma granulite forming event in East Gondwana and the observation that the apparent polar wander path for Gondwana does not form a spatially and temporally coherent pattern until roughly the same time. The Kuunga orogeny may have resulted from the collision between Australia and Antarctica with the rest of Gondwana.

Magnetic iron oxides in a sequence of pillow basalts that were dredged from the Atlantic Ocean floor have been studied to characterize titanomaghemite and to define the processes of maghemitization. Distances from the spreading ridge and ages (in parentheses) of the samples are 0-10 (0-1), 160 (9), 450 (26), and 900 km (70 Ma).Iron titanium oxides occur as 1 to 10 mu m-sized dendritic and cruciform-shaped crystals with identical appearances in all samples and with no signs of change or significant heterogeneity in composition or structure as observed by TEM and AEM. Parameters change progressively from the youngest to the oldest, e.g., Curie temperature = 180 to 360 degrees C; lattice parameter = 8.466 to 8.361 Angstrom; number of octahedral cations per cell from Rietveld refinement = 14.8 to 12.1; mean hyperfine (internal) fields at 300 K from Mossbauer data = 37 to 45 T. The large Ti contents (Uv(60) to Uv(70)) are nearly constant. SAED patterns show superstructure reflections only for the oldest sample.The youngest sample has parameters corresponding to nearly unoxidized titanomagnetite, whereas the oldest is near-end-member titanomaghemite. Intermediate samples are partially altered but display no superstructure reflections, implying a lack of significant ordering of vacancies. The data therefore show that the process of (titano)maghemitization has two distinctly different components: (1) oxidation and loss of Fe, with creation of disordered vacancies, and (2) ordering of vacancies. The data collectively imply a process dominated by solid state diffusion of Fe from the crystals, oxidation of Fe, and creation of vacancies wherein the O closest-packed framework is preserved, in sharp contrast to a model of addition of O or to dissolution and neocrystallization.

During the Neoproterozoic and Palaeozoic the two continents of Baltica and Laurentia witnessed the break-up of one supercontinent, Rodinia, and the formation of another, but less long-lived, Pangea. Baltica and Laurentia played central roles in a tectonic menage a trois that included major orogenic events, a redistribution of palaeogeography and a brief involvement of both with Gondwana. Many of these plate re-organisations took place over a short time interval and invite a re-evaluation of earlier geodynamic models which limited the speeds at which large continental plates could move to an arbitrarily low value.Baltica and Laurentia probably shared a common drift history for the time interval 750-600 Ma as they rotated clockwise and drifted southward from an equatorial position during the opening of the Proto-Pacific between Laurentia and East Gondwana (initial break-up of Rodinia). On their combined approach toward the south pole, Baltica and Laurentia were glaciated during the Varanger glaciations. Although the two continents drifted toward the south pole during the Late Proterozoic, they began to separate at around 600 Ma (rift to drift) to form the Iapetus Ocean through asymmetric rifting and relative rotations of up to 180 degrees. Initiation of rifting on the Baltic margin is marked by the 650 Ma Egersund tholeiitic dykes (SW Norway) which contain abundant lower crustal zenoliths, and the tholeiitic magma was probably derived from a mantle plume.In latest Precambrian time, the final redistribution of Rodinia is characterised by high plate velocities. In particular, Laurentia began a rapid, up to 20 cm/yr, ascent to equatorial latitudes and essentially stayed in low latitudes throughout most of the Palaeozoic. The high velocities suggest either that Laurentia was pushed off a lower mantle heat anomaly originating from supercontinental mantle insulation or that Laurentia was pulled toward a subduction-generated cold spot in the proto-Pacific. Baltica, except for a short and rapid excursion to lower latitudes in the Late Vendian, remained mostly in intermediate to high southerly latitudes and closer to the Gondwana mar in until Early Ordovician times.In Early Ordovician times, Arenig-Llanvirn platform trilobites show a broad distinction between the continents of Laurentia/Siberia/North China Block (Bathyurid), Baltica (Ptychopygine/Megalaspid) and the areas of NW Gondwana/Avalonia/Armorica (Calymenacean-Dalmanitacean). During the Ordovician, Baltica rotated and moved northward, approaching close enough to Laurentia by the late Caradoc for trilobite and brachiopod spat to cross the intervening Iapetus Ocean. Ducking appears to have been irregular both in time and manner: the collision between Scotland/Greenland and western Norway resulted in the early Scandian Orogeny in the Silurian (c. 425 Ma), but further south, there is evidence of late Silurian impingement with subduction of Avalonian continental crust (in England and Ireland) below the eastern edge of Laurentia until the Emsian. In the northern Appalachians the main time of collision appears to have been during the Emsian/Eifellian Acadian Orogeny,Recent analyses invalidates the traditional concept of a sustained orthogonal relationship between Baltica and Laurentia across a single Iapetus Ocean throughout the Caledonide evolution. The active margin of Baltica (Scandinavian Caledonides) faced Siberia during the Late Cambrian and Early Ordovician with oceanic separation between these landmasses in the order of 1200-1500 km. This may explain the local occurrences of Siberia-Laurentian type Bathyarid trilobite faunas in Central Norwegian Caledonian nappes, earlier interpreted as Laurentia-Baltica trilobite mixing. Subsequent counterclockwise rotation of Baltica transferred the Caledonian margin in the direction of Laurentia by Silurian times, when the two continents once again started to collide to form Euramerica. This rotation, along with the strongly asymmetric opening of the Iapetus at around 600 Ma, demonstrates a complexity in Precambrian-Palaeozoic plate tectonics, i.e. a collage of metastable plate boundaries which have perhaps too often been simplified to an orthogonal Wilson cycle tectonic scenario.

We report palaeomagnetic data and a composite magnetic polarity sequence for Middle and Upper Triassic rocks assigned to the Anton Chico Member of the Moenkopi Formation and Chinle Group, respectively, exposed along the eastern flank of the Sangre de Cristo Mountains and in the Tucumcari Basin of eastern and northeastern New Mexico. Thermal demagnetization isolates a well-defined, dual polarity, characteristic magnetization, carried in most cases by haematite and interpreted as an early acquired chemical remanent magnetization (CRM). Characteristic magnetizations from 74 palaeomagnetic sites (one site = one bed) are used to define a magnetic polarity sequence, which we correlate with previously published Triassic data obtained from both marine and non-marine rocks. Preliminary correlation suggests that the resolution of magnetostratigraphic data derived from continental strata is not necessarily of lesser quality than that from marine rocks. On the basis of the magnetostratigraphic data, a profound unconformity is believed to separate lower-middle Norian and upper Norian-Rhaetian strata of the Chinle Group. Palaeomagnetic poles derived from selected sites in steeply dipping (> 85 degrees) strata for the Middle Triassic (Anisian, similar to 240 Ma: 50 degrees N 121 degrees E; N = 8), late Carnian-early Norian (similar to 225 Ma: 53 degrees N 104 degrees E; N = 16), and late Norian-Rhaetian (similar to 208 Ma: 59 degrees N 77 degrees E; N = 8) are in relatively good agreement with previously published data for the Moenkopi Formation and Chinle Group and related strata in southwest North America. None the less, comparison with palaeomagnetic poles obtained from gently dipping or flat-lying Triassic strata from this study (Anisian, 46 degrees N 112 degrees E; N = 13; late Carnian, 54 degrees N 87 degrees E; N = 12) and previously published Triassic poles in southwest North America suggest that a modest 'apparent rotation' not greater than about 5 degrees affects declinations from steeply dipping rocks. The distribution of palaeomagnetic poles indicates similar to 25 degrees (angular distance) of apparent polar wander between about 240 and 208 Ma.

Paleomagnetic analyses were completed on two volcanic units of the Exploits Group in Newfoundland's Central Mobile Belt, which are part of an Ordovician arc-back-arc system. The Tea Arm Volcanics display scattered and unstable characteristic directions that cannot be interpreted. However, stable end-points in six sites supported by great-circle analysis of seven sites in the mid-Arenigian to Llanvirnian Lawrence Head Volcanics yield a tilt- and strike-corrected characteristic direction of D = 56°, I = 23°, (α95 = 19°), k = 14, N = 6). The magnetization is carried by magnetite and passes a tilt test. The corresponding paleolatitude of 12° ± 10° is interpreted as southerly, and is similar to the paleolatitude of 11° ± 4° for arc volcanics of the nearby Moreton's Harbour Group. However, this near-Laurentian paleolatitude is distinctly different from the paleolatitude of 31° ± 8° reported for the Robert's Arm-(Cottrell's Cove)-Chanceport-Summerford volcanic terrane. The low paleolatitude of the Lawrence Head Volcanics supports the location of a major Ordovician subduction system near the Laurentian margin of Iapetus, whereas the present-day juxtaposition of latitudinally distinct elements in the area represents a complex accretionary history involving Early Silurian or older thrusting and younger strike-slip faulting of several Iapetan island-arc terranes to Laurentia.

A palaeomagnetic study of Permo-Triassic cover rocks in the Asturian-Cantabrian Are has revealed characteristic magnetizations in 136 samples from 14 sites, yielding a mean palaeopole for each of the three regions studied (Cabo de Penas, Villaviciosa, and Tudanca). Previous work on remagnetized Palaeozoic formations in the area revealed Permian (or younger) tightening of the Are on the basis of palaeomagnetic declinations that show relative rotations of more than 100 degrees. According to the new palaeopoles obtained in this study, these rotations are not present to any comparable extent in the Permo-Triassic deposits, allowing us to conclude that the bulk of the rotations in the Are is Permian in age.

The 810 Ma Gagwe-Kabuye lavas and the 743 Ma Mbozi gabbro-syenite complex of the Congo Craton in East Africa were sampled for paleomagnetic study in an effort to test a variety of tectonic models proposed for Neoproterozoic times. The paleomagnetic pole obtained from the Gagwe-Kabuye lavas falls at 25°S, 273°E (δp = 7°, δm = 12°) and compares favorably to a previously published paleomagnetic pole obtained from these rocks. The Mbozi complex pole yields a paleomagnetic pole at 46°N, 325°E (δp = 5°, δm = 9°) and differs significantly from a previously determined pole for the Mbozi complex. A comparison of these paleomagnetic poles to Laurentian poles of the same age suggests that the Congo Craton may not have constituted part of the Rodinia supercontinent in the configuration proposed by Dalziel (1992). An analysis of reliable paleomagnetic poles from the Gondwana blocks for the interval from 810 to 510 Ma reveals a coherent swathe of poles from 550 to 510 Ma and a scatter of pre-600 Ma poles. Our interpretation of the available paleomagnetic and tectonic data for this interval is consistent with the formation of Gondwana by two distinct orogenic events. This assembly resulted in the East Africa Orogen between 800 and 650 Ma and a younger Kuunga Orogen at 550 Ma outboard of the East Africa Orogen with possible sutures located in Sri Lanka, southern India and Enderby Land (East Antarctic Craton).

We report paleomagnetic data for 26 accepted sites collected in two sections of flat‐lying strata of the upper Carnian‐lower Norian (∼225 Ma) Dockum Group, northwest Texas. Six additional sites in coarse‐grained conglomeratic sandstones gave no usable results. The total assemblage of 26 VGPs is streaked along the Late Triassic ‐ earliest Jurassic track of the North America apparent polar wander path and their mean is inconsistent with the accepted upper Carnian‐lower Norian reference pole. In detail, 12 sites in grayish white (nonhematitic) sandstones have weak magnetizations (less than about 1 mA/m) carried by magnetite or maghemite that give a paleopole at 56.4°N–96.3°E (N=12 dual‐polarity VGPs; K=44.2; A95=6.6°) in close agreement with other results for upper Carnian‐lower Norian rocks in North America. The 14 remaining sites in tan and redcolored (hematitic) sandstones, siltstones, and claystones give high unblocking temperature characteristic magnetizations carried by hematite, with paleopoles at 59.0°N–53.8°E (normal; N=7 VGPs, K=62.4, A95=7.6°) and 59.3°N–77.8°E (reverse; N=7 VGPs, K=204.2, A95=4.2°). These poles fall along the younger track of poles and near the J‐I cusp of the North American APWP as defined by (unrotated) poles derived from Colorado plateau rocks. We suggest that the characteristic magnetization of the non‐hematitic sandstones is an “early” magnetization, acquired during or soon after deposition. However, the characteristic magnetization of the red bed sites is interpreted as a secondary magnetization, for which we infer an earliest Jurassic age. The secondary origin for this magnetization is supported by the observation of conflicting magnetostratigraphies. The extreme westward position of the poles derived from red bed sites, particularly those with normal polarity, confirms the general position of the J‐I cusp indicated by poles in the Piedmont province, the Newark basin, and the Colorado plateau; it also suggests that the magnitude of rotation of the Colorado plateau is no greater than about 5°. We compile an apparent polar wander path for North America, including Colorado plateau data, which suggests a fast rate of apparent polar wander throughout the Triassic period (about 0.8°/m.y.) with a gradual increase that preceded the opening of the Atlantic.

In many mobile belts, paleomagnetic directions from elastic sedimentary rocks are typically shallower than directions from associated igneous rocks. This discordance raises questions about the reliability of sedimentary rocks and can lead to controversial paleogeographic reconstructions. In Newfoundland, the controversy over Silurian paleogeography arises from a difference between results from elastic redbeds (with anomalously shallow directions that place Newfoundland at the paleoequator in the Silurian) and coeval epicontinental volcanics (with steeper directions that place Newfoundland at a more southerly paleolatitude). This study tests the possibility that the anomalously shallow redbed directions are the result of internal strain or inclination error related to deposition. We specifically compare coarse- and fine-grained lithologies under the assumption that such remanence-altering effects will be different in rocks with differing mechanical competence.We show that inclinations correlate with lithologic variations, but in a manner opposite to that predicted for the accumulation of internal strain, Based upon a strict set of demagnetization criteria, clearly defined single-component characteristic directions are only observed in the finer grained sandstones. These well-determined directions yield a pre-folding and pre-rotational dual polarity magnetization that places Newfoundland at a paleolatitude of 23 degrees (S) +/- 9 degrees in the Silurian. This revised paleolatitude is consistent with the volcanic results and with the paleolatitude predicted for Newfoundland from the North American reference path. In contrast, coarse-grained sandstones yield shallower directions that appear to be composites of opposite polarity magnetizations or weakly defined directions with strongly contaminating overprints. Because the shallow magnetizations in the coarser grained samples may have originated as a depositional remanence, we also suspect inclination error related to deposition. The degree of inclination error is consistent with values predicted for redbeds from previously published experimental results. These observations suggest that grain-size considerations are important in assessing the reliability of redbed magnetizations. Recognition of the relationship between grain size and remanence acquisition allows us to resolve the Silurian paleolatitude controversy for Newfoundland.

Two characteristic magnetic components (A1 and A2) with distinctive non-overlapping bimodal distributions were palaeomagnetically isolated from what has long been considered as one single phase of post-Oligocene pre-late Miocene basaltic volcanicity in the 'east- and west-of-Cairo areas'. A1 yields a mean Dec/Inc=198 degrees/-24 degrees, alpha 95=2.6 degrees and K=417 based on the site-means in tilt-corrected co-ordinates with a corresponding palaeomagnetic north pole at 66 degrees N/167 degrees E and A95=2.3 degrees using site-mean pole averaging. A2 yields a distinct mean Dec/Inc=197 degrees/-51 degrees with alpha 95=3.1 degrees and K=270 corresponding to a palaeomagnetic north pole at 76 degrees N/111 degrees E and A95=3 degrees. A1 and A2 both pass the tilt test; the two components are never recorded at the same site and no intermediate direction was observed.The bi-modal grouping of the A1-A2 directions and their non-overlapping cones of 95% confidence indicate that the group means of Al and A2 are significantly different in tilt-corrected coordinates at the 95% level of confidence and represent two different ages of magnetization. Therefore, we argue that two distinct basaltic episodes occurred, which were not previously distinguished in the areas 'east- and west-of-Cairo'. Ar-39/Ar-40 whole rock analyses appear to yield Early Miocene ages, but there are too many problems with the argon release spectra to rely on the integrated ages for more precise temporal definition.A petrological study subsequently carried out on the two distinct basaltic episodes revealed that the A1 episode is represented by very fine-grained porphyritic olivine-bearing basalts with dust-like groundmass rich in opaques made up of homogeneous magnetite grains that are rarely intergrown with homogeneous ilmenite lamellae. In contrast, the A2 episode consists of medium- to coarse-grained holocrystalline non-porphyritic doleritic basalts with inhomogeneous magnetite and ilmenite grains that show a variety of intergrowths.The two poles of this study are dearly virtual geomagnetic poles because of imperfect averaging of the secular variation of the geomagnetic field due to the limited number of exposures and the rapid cooling of the Bows. Even so, a comparison with the APWP of Africa shows that the segment joining the two poles of the present study has a similar pattern as that of the APWP derived from DSDP sediments of the Atlantic part of Africa or those rotated from the North American craton and stable Europe and allows the authors to calibrate part of this APWP as Early. Miocene. The available palaeomagnetic results imply a northward movement of Africa during the Tertiary followed by a clockwise rotation with respect to the axial geocentric dipole.

The Nyanzian System lavas of western Kenya are believed to be the oldest rocks of the Tanzanian Craton. Intrusive age relationships suggest an age ⩾2850 Ma although direct attempts at dating the Nyanzian have produced disparate results. Our study involves a suite of samples collected from the Nyanzian basalts, pillow basalts, andesites and rhyolites from sixteen sites in western Kenya. These rocks yield a tilt-corrected paleomagnetic pole at 14°N, 150°E (K=59, dp=5°, dm=7°). This pole is constrained to be older than the first (D1) deformation (>2472±30 Ma) by positive fold, conglomerate and reversal tests. Analysis of the paleomagnetic data base for three African cratonic nuclei (Tanzanian, Kaapvaal/Zimbabwe and West Africa) for the time period from 2.0 Ga to 3.0 Ga demonstrates a paucity of well-dated poles, although there are several poles from the Kaapvaal/Zimbabwe and Tanzanian Cratons which allow “spot-readings” of their relative positions. We demonstrate, based on these data, that the Kaapvaal/Zimbabwe and Tanzanian Cratons were drifting independently at ∼ 2875 Ma, ∼ 2700 Ma and ∼ 2450 Ma. This independent motion of the Tanzanian and Kaapvaal/Zimbabwe Cratons indicates that previously proposed models involving African cratonic coherence can no longer be considered valid for the time period from 2850 to 2500 Ma.

The Neoproterozoic interval (1000–540 Ma) contains ample evidence for a series of glacial intervals. These include the 750–700 Ma Sturtian glaciation, the 625–580 Ma Marinoan-Vendian glaciation and the 600–550 Ma Sinian glaciation. Paleomagnetic evidence has suggested that many of these glaciations occurred at tropical latitudes ( ⩽ 25°) and this led to a number of theories that attempt to explain the occurrence of these anomalously low latitude glaciations (e.g., an increase in the axial tilt of the earth, an equatorial low-orbit ice-ring, rapid equator to pole continental drift, incorrect identification of impact deposits as glacial deposits or secondary magnetizations misidentified as primary). New paleomagnetic data for Laurentia, China, Baltica and parts of Gondwana are combined with a reanalysis of previously published data to demonstrate that the Neoproterozoic glaciations may well all have occurred above 25° latitude. Climate models using a juvenile Sun of slightly lower luminosity, lower CO2 levels and coupling to Milankovitch cycles suggest that ice sheets could extend to within ±25° of the Neoproterozoic equator. Thus, the new paleomagnetic data and climate models offer an alternative explanation for the Neoproterozoic glaciations that is consistent with the waxing and waning of intermediate latitude ice sheets to form the conformable sequences of warm climate-cold climate strata.

Devonian carbonates in the Cantabrian Arc reveal characteristic magnetizations with coherent, shallow, upward inclinations. The magnetizations appear to be carried by magnetite. Within‐site directions are very well grouped, but site‐mean declinations range from easterly to south–southwesterly in in situ as well as tilt‐corrected coordinates, as has also been observed in previous studies of other formations in the arc. The widely varying declinations of all studies roughly correlate with the overall structural trends of the arc and suggest that the sites underwent rotations in a process that involved folding about vertical axes and tightening of arc. Upon tilt correction the inclinations of our study, on the other hand, become scattered, and it is concluded that the magnetizations were acquired after Late Carboniferous folding about horizontal axes. The oroclinal rotations therefore also must have occurred after the Late Carboniferous folding phase. The previous paleomagnetic results had been interpreted mostly as primary magnetizations residing in hematite. However, inclination only fold‐tilt tests applied to these results suggest that many, if not all, of the directions were acquired during the earlier stages of the Late Carboniferous folding. Thus all paleomagnetic results from the Cantabrian Arc appear to be remagnetizations, but the ages of the remagnetizations vary from pre‐to synfolding for the mostly hematitic formations to postfolding for the Devonian carbonates. The reversed‐polarity inclinations of the hematite‐bearing formations have mean values ranging from +25° to +5°, whereas the carbonates have a mean inclination of −8°. On the basis of inclinations predicted for the area from results from stable Europe, the Pyrenees, and the Iberian Meseta, the ages of these remagnetizations can be inferred to range from about 320 Ma to 260 Ma. Because all the remagnetizations reveal rotated declination patterns, the oroclinal rotations occurred well after the main phase of Hercynian deformation (320–280 Ma). While the timing of the rotations is unconstrained at the younger end, they must have occurred during or after the Permian (best estimate is less than 260 Ma), which is much later than anticipated from other geological considerations.

The Vendian/Cambrian segment of the Lauretian apparent polar wander path (APWP) has been poorly constrained and the subject of some controversy. The Catoctin volcanic province in central Virginia is well‐dated at 570±35 Ma (Rb‐Sr) and 597±18 Ma (U‐Pb) and therefore presented an excellent paleomagnetic target for resolving the Laurentian Vendian‐Cambrian APWP. A total of 206 samples from 32 sites were collected from the Catoctin basalts, feeder dikes and sills. The study revealed three ancient directions of magnetization. The youngest, C component, fails the fold test and yielded a characteristic in situ direction of D = 147°, I = +44° (k = 21, α95 = 9°). The corresponding paleopole falls along the Middle Ordovician segment of the Laurentian APWP and we consider this component to be the result of a Taconic remagnetization. The second component, the B component, is carried by hematite, exhibits dual‐polarities and passes a fold test. The tilt‐corrected B component characteristic direction is D = 92°, I = +17° (k = 16, α95 = 13°). The corresponding paleopole at 4°S, 193°E falls near a well‐established Late Cambrian (505 Ma) pole for Laurentia, and we consider this component to be a remagnetization during a Late Cambrian tectonic event in the central Appalachians. The third component isolated in the Catoctin basalts, the A component, yields a tilt‐corrected mean of D = 68°, I = +84° (k = 59, α95 = 9°). This component passes a fold and reversal test. A suite of samples was collected from two Catoctin feeder dikes and surrounding country rocks that yield a positive baked contact test. The A pole at 43°S, 128°E falls significantly away from previously proposed Vendian poles for Laurentia. A reevaluation of previous paleomagnetic studies from coeval rock units reveals similarly steep directions and leads us to propose a new APWP. This new APW track indicates that Laurentia was located near the pole during the interval 615–580 Ma and drifted rapidly (16 cm yr−1) toward its Late Cambrian equatorial position.