Palaeomagnetic evidence from the Ordovician and Silurian of northwest Peninsular Malaysia

The Ordovician and Silurian Setul Limestone of the Langkawi Islands, northwest Peninsular Malaysia, has a mean magnetic vector ofD = 338°,I = 62° after cleaning and correction for tilt. This is equivalent to a palaeolatitude of 43°, and a palaeomagnetic pole at 46°N, 76°E. The Silurian part of the Setul limestone also shows a similar direction. The Ordovician results are equivalent to a palaeolatitude of 43°, N or S. Recent reconstructions, based on palaeontology, place Indochina and China in the northern hemisphere in the Ordovician; if this is correct, a palaeolatitide of 43° for Langkawi would imply that Malaya-Indochina was the most northerly continental fragment at that time.     

Palaeomagnetism of the Msissi norite (Morocco) and the Palaeozoic reconstruction of Gondwanaland

Palaeomagnetic results are reported from eight sites in an Upper Devonian basic intrusion (the Msissi norite) in southeast Morocco. Specimens from one site are suspected of having been affected by lightning, but results from the other seven sites indicate the presence of a less-stable component, probably of viscous origin. The pole position corresponding to the stable component(0.5°S, 25°E, A₉₅ = 16.5) is interposed between the Middle Cambrian/Ordovician pole and the Lower Carboniferous pole on the African polar wander curve. When the southern continents are reassembled on the Smith/Hallam reconstruction of Gondwanaland the new Moroccan Devoniån pole is in excellent agreement with the corresponding portion of the main Australian polar-wander curve. This places additional constraints on the possible date of fusion of the separate Southeast Australian plate with the rest of Gondwanaland, postulated recently on palaeomagnetic grounds by M.W. McElhinny and B.J.J. Embleton (1974). The combined African/Australian polar-wander curve is compared with the South American curve, and two possible interpretations of available data are discussed, one involving possible relative tectonic motion between South America and the rest of Gondwanaland during the Lower and Middle Palaeozoic, and the other, favoured here, requiring a reassessment of the ages of several South American pole determinations.     

Palaeomagnetism and K/Ar results from the Duncansby volcanic neck,NE Scotland: superimposed magnetizations,age of igneous activity,and tectonic implications

The Duncansby volcanic neck, intruding the Middle Devonian red beds of north Caithness, Scotland, has revealed two significantly different axes of magnetization, yielding pole positions at 149°E, 24°N and 126.5°E, 60°N, respectively. The first pole, which is interpreted as corresponding to the oldest magnetization, is in perfect agreement with Devonian polar estimates from west of the Great Glen Fault. It is tentatively suggested therefore that the Duncansby neck correlates with the Late Devonian volcanism in the nearby Orkney Islands though palaeomagnetism allows an upper age estimate of around Middle Carboniferous. The data support an earlier proposition of there being a palaeomagnetic discordance across the Great Glen Fault that can be interpreted in terms of a large-scale late- or post-Devonian transcurrent movement along this fracture zone. The original (? Late Devonian) magnetization has been nearly completely erased by the second phase of magnetization which, according to its pole position, most likely dates from about the Middle Jurassic. The latter magnetization is thought to be a consequence of burial, the coastal districts of Caithness having participated in the general subsidence of the North Sea area in late Palaeozoic and Mesozoic times. The burial magnetization, involving VPTRM and or TCRM processes, is considered to have been “frozen-in” as a result of uplift in connection with the well-documented mid-Jurassic tectonic phase that affected the northern North Sea basin, including the adjacent Moray Firth area. K/Ar analyses of the Duncansby intrusion give apparent ages ranging from 258 to 239 Ma. These dates, which lie between the two geological events inferred from palaeomagnetism, are not seen as true rock ages but rather as the result of a partial Ar loss during burial reheating.     

Palaeomagnetism and K-Ar ages of triassic igneous rocks from the Ischigualasto-Ischichuca Basin and Puesto Viejo Formation,Argentina

The palaeomagnetism of Middle Triassic (224 ± 5 m.y.) igneous rocks from the Ischigualasto-Ischichuca Basin (67°40′W, 30°20′S) was investigated through 86 oriented hand samples from 11 sites. At least one reversal of the geomagnetic field has been found in these rocks. Nine sites yield a palaeomagnetic pole at 239°E, 79°S (α₉₅ = 15°, k = 13).The K-Ar age determinations of five igneous units of the Puesto Viejo Formation give a mean age of 232 ± 4 m.y. (Early Triassic). The palaeomagnetism of six igneous units of the Puesto Viejo Formation (68°W, 35°S) was investigated through 60 oriented samples. These units, two reversed relative to the present magnetic field of the Earth and four normal, yield a pole at 236°E, 76°S (α₉₅ = 18°, k = 14).Data from the Puesto Viejo Formation indicate, for the first time on the basis of palaeomagnetic and radiometric data, that the Illawarra Zone, which defines the end of the Kiaman Magnetic Interval, extends at least down to 232 ± 4 m.y. within the Early Triassic. The palaeomagnetic poles for the igneous rocks of the Ischigualasto-Ischichuca Basin and Puesto Viejo Formation form an “age group” with the South American Triassic palaeomagnetic poles (mean pole position: 239°E, 77°S; α₉₅ = 6.6°, k = 190). The Middle and Upper Permian, Triassic and Middle Jurassic palaeomagnetic poles for South America would define a “time group” reflecting a quasi-static interval (mean pole position: 232°E, 81°S; α₉₅ = 4°, k = 131).     

Rotation of the Iberian arc: Palaeomagnetic results from North Spain

NRM directions measured from 32 sites in Middle Cambrian, Upper Silurian/Lower Devonian and Lower Carboniferous redbeds follow the trend of the Variscan arc in North Spain. Thermal demagnetisation does not significantly alter this pattern. Fold tests show that the NRM is earlier than the ?₁ folds which form the arc; consistency of angle between bedding and the tilt-corrected NRM inclination (22–28°), similarity of the corresponding palaeolatitudes to Carboniferous values and microscopic evidence of Variscan redistribution of hematite indicates that the magnetisation is post-Lower Carboniferous. A statistical plot of the orientation of ?₁ fold traces against angle between ?₁ fold traces and declination of NRM shows that where these folds curve through 165° the NRM has been rotated through 110°: the arc is an orocline. Restoration of this rotation, and that needed to close the Bay of Biscay, brings the calculated mean palaeomagnetic pole reasonably close to the Upper Carboniferous part of the apparent polar wander path for Europe.     

Lower Jurassic palaeomagnetic results from Yorkshire,England, and their implications

Palaeomagnetic study of Middle Liassic siltstones shows a stable magnetization with a mean direction of D = 12.3°, I = 64.6° (N = 60, k = 26, α₉₅ = 3.9°) corresponding to a palaeomagnetic pole at 79.8°N, 125.6°E, similar to that for southern Germany and confirming predictions based on palaeogeographic reconstructions using North American data. Sideritic concretions of Lower Liassic age show a higher magnetic stability with a mean direction of D = 12.6°, I = 61.4° (N = 125, k = 50, α₉₅ = 1.8°) which is not significantly different from the siltstones. This confirms the sedimentological evidence that suggests that such concretions grew very shortly after deposition, i.e. within the Liassic, and suggests that similar concretions of other ages could thus be used for palaeomagnetic studies. Although the Liassic palaeomagnetic pole (76.9°N, 134.7°E), based on this work, appears valid it is still not possible to evaluate a sensible Mesozoic polar wandering curve for the North Atlantic bordering continents.     

Paleomagnetism of the Lower Ordovician and Cambrian sedimentary rocks in the section of the Narva River right bank: For the construction of the Baltic Kinematic model in the Early Paleozoic

The paleomagnetic study of the Lower Ordovician and Cambrian sedimentary rocks exposed on the Narva River’s right bank revealed a multicomponent composition of natural remanent magnetization. Among four distinguished medium- and high-temperature magnetization components, the bipolar component, which carries the reversal test, is probably the primary component and reflects the geomagnetic field direction and variations during the Late Cambrian and Early Ordovician. The pole positions corresponding to this component have coordinates 22°N, 87°E (dp/dm = 5°/6°) for the Late Cambrian, and 18°N, 55°E (dp/dm = 5°/7°) for the Early Ordovician (Tremadocian and Arenigian). Together with the recently published paleomagnetic poles for the sections of the Early Ordovician in the Leningrad Region and the series of poles obtained when the Ordovician limestones were studied in Sweden, these poles form new key frameworks for the Upper Cambrian-Middle Ordovician segment of the apparent polar-wander path (APWP) for the Baltica. Based on these data, we propose a renewed version of the APWP segment: the model of the Baltica motion as its clockwise turn by 68° around the remote Euler pole. This motion around the great circle describes (with an error of A₉₅ = 10°) both variations in the Baltic position from 500 to 456 Ma ago in paleolatitude and its turn relative to paleomeridians. According to the monopolar components of natural remanent magnetization detected in the Narva rocks, the South Pole positions are 2°S, 351°E (dp/dm = 5°/9°), 39°S, 327°E, (dp/dm = 4°/7°), and 42°S and 311°E (dp/dm = 9°/13°). It is assumed that these components reflect regional remagnetization events in the Silurian, Late Permian, and Triassic.     

Paleomagnetism of the western Cape Fold belt, South Africa, and its bearing on the Paleozoic apparent polar wander path for Gondwana

In order to test two different proposals for the poorly defined African Paleozoic apparent polar wander path (APWP), a paleomagnetic study was carried out on Ordovician through Carboniferous clastic sediments from the Cape Fold belt, west of the 22nd meridian. One proposal involves a relatively simple APWP connecting the Ordovician Gondwana poles in North Africa with the Late Paleozoic poles to the east of South Africa in a more or less straight line crossing the present equator in the Devonian. The other proposal adds a loop to this path, connecting Ordovician poles in North Africa with poles to the southwest of South Africa and then returning to central Africa. This loop would occur mainly in Silurian time. New results reported herein yield paleopoles in northern and central Africa for Ordovician to lowermost Silurian and Lower to Middle Devonian formations. The best determined paleopole of our study is for the Early Ordovician Graafwater Formation and falls at 28°N, 14°E (k = 25, α₉₅ = 8.8°, N = 28 samples). The other paleopoles are not based on sufficient numbers of samples, but can help to constrain the apparent polar wander path for Gondwana. Our results give only paleopoles well to the north of South Africa and we observe no directions within the proposed loop. Hence, if the loop is real, it must have been of relatively short duration (60–70 Ma) and be essentially of Silurian/Early Devonian age, implying very high drift velocities for Gondwana (with respect to the pole) during that interval.     

Palaeomagnetism of middle Ordovician volcanic rocks from Quebec

This palaeomagnetic study is centered on agglomerates and volcanic rocks from the western margin of the Appalachian belt in the Drummondville-Actonvale-Granby area, Quebec (long.: 72°30′W, lat.: 46°00′N). It involves a total of 36 oriented samples (111 speciments) distributed over eleven sites. Both thermal and AF cleaning techniques were used to isolate residual remanent components. The dispersion of the directions is slightly reduced after AF cleaning and thermal treatment.The palaeopole position obtained is 191°E, 6°N (d_m = 14°, d_p = 7°) after thermal treatment and 164°E, 19°N (d_m = 11°, d_p = 6°) after AF cleaning. The polarity of most of the sites (two exceptions) are reversed. The thermal-treated data appear to be relatively stable and an approximate value of the primary magnetization is extracted from them. The palaeopole obtained does not lie close to the tentatively defined position of the Cambrian and Ordovician poles from rocks of the North American plate; it is located near the Upper Cambrian and Lower Ordovician poles from eastern Newfoundland and the Lower Ordovician pole from the Caledonides in Europe.     

The palaeoposition of Madagascar: Remanence and magnetic properties of late Palaeozoic sediments

Palaeomagnetic results are reported from the predominantly green sediments of the Upper Permian to Lower Triassic Sakamena Group and the Upper Carboniferous to Lower Permian Sakoa Group of Madagascar. Secondary magnetizations could only be removed successfully through thermal demagnetization procedures and then only if the cleaning process was completed by 450°C. Heating in air caused extensive magnetochemical changes to occur above this temperature. Coercivity spectrum analysis and low-temperature characteristics of the heated and unheated green sediments indicate that considerable amounts of fine-grained single-domain magnetite are formed at 500°C or more from some non-magnetic mineral, probably the iron silicates. For this reason consistent palaeomagnetic data could only be obtained from about half the samples collected. Results from 4 sites (19 samples) of the Lower Sakamena Group give a palaeomagnetic pole at 64.9S, 113.9E (A₉₅ = 5.6°) and 3 sites (16 samples) from the Glacial Series of the Sakoa Group give a pole at 47.9S, 84.1E (A₉₅ = 8.1°). When compared with corresponding data from Africa these results confirm and strengthen our previous conclusions from the Triassic-Jurassic Isalo Group regarding the palaeoposition of Madagascar. All three poles are only consistent with the Smith and Hallam reconstruction which places Madagascar off the eastern coast of Africa adjacent to Kenya and Tanzania.     

Palaeomagnetic directions in Precambrian basic intrusives of the Mount Isa Province,Australia

Palaeomagnetic investigation of basic intrusives in the Proterozoic Mount Isa Province yields three groups of directions of stable components of NRM after magnetic cleaning in fields up to 50 mT (1 mT= 10 Oe). The youngest group (IA) includes results from the Lakeview Dolerite, and yields a palaeomagnetic pole at 12°S, 124°E (A₉₅ = 11°). The second group (IB) has a palaeomagnetic pole 53°S, 102°E (A₉₅ = 11°). The third group (IC) is derived from the Lunch Creek Gabbro and contains normal and reversed polarities of magnetization with a palaeomagnetic pole at 63°S, 201°E (A₉₅ = 9°). Some samples from the gabbro have anomalously low intensities of remanent magnetization in obscure directions attributed to the relative enhancement of the non-dipole component of the palaeomagnetic field during polarity reversal. The present attitude of the igneous lamination is probably of primary, not tectonic origin.     

Palaeomagnetism of the dyke swarms of the Gardar Igneous Province,south Greenland

In the western part of the Gardar Igneous Province of southern Greenland, lamprophyre dykes intruded at ca. 1276-1254 m.y. RbSr biotite ages yield a palaeomagnetic pole at 206.5°E,3°N (nine sites, dψ = 5.1°, d_χ = 10.1°) Slightly younger dolerite dykes with RbSr biotite ages in the range 1278-1263 m.y. give a pole at 201.5°E,8.5°N (24 sites, dψ = 4.7°, d_χ = 9.4°), and the syeno-gabbro ring dyke of the Kûngnât complex (RbSr isochron age 1245 ± 17 m.y.) cutting both of these dykes swarms, gives a pole at 198.5°E, 3.5°N (four sites, dψ = 2.3°,d_χ = 4.4°). All these rock units have the same polarity and the poles are identical to those from Mackenzie and related igneous rocks of North America (1280-1220 m.y.) after closure of the Davis Strait; they confirm that this part of the Gardar Province is a lateral extension of the Mackenzie igneous episode within the Laurentian craton.In the Tugtutôq region of the eastern part of the Gardar Province 47 NNE-trending dykes of various petrologic types, and intruded between 1175 ± 9 and 1168 ± 37 m.y. (RbSr isochron ages) yield a palaeomagnetic pole at 223.9° E, 36.4°N (dψ = 4.1°, d_χ = 6.1°). Fifteen other dykes in this swarm were intruded during a transitional phase of the magnetic field which, however, does not appear to have achieved a complete reversal over a period of several millions of years. The majority of dykes studied are highly stable to AF and thermal demagnetisation and contain single high blocking temperature components with single Curie points in the range 380–560°C.Palaeomagnetic poles from the Gardar Province between ca. 1330 and 1160 m.y. in age define the earlier part of the Great Logan apparent polar-wander loop; they correlate closely with contemporaneous North American results and confirm the coherence of the Laurentian craton in Upper Proterozoic times.     

Palaeomagnetic study of the Swedish rapakivi suite: Proterozoic tectonics of the Baltic Shield

The major Proterozoic igneous intrusions in the Swedish sector of the Baltic Shield are the Ragunda complex (1293 m.y., palaeomagnetic pole 165°E, 54°N) and the Nordingrågabbro-granite-anorthosite complex (1385 ± 30 m.y.). The latter body has been partially remagnetised by later post-Jotnian dolerites (1254 m.y.), and sites influenced by the dolerites have a stable magnetisation with a mean direction D = 45°, I = ?39°, (α₉₅ = 4.3°). Elsewhere, the gabbro-anorthosite facies have a magnetisation of dual polarity predating the dolerite and recoverable at various stages of thermal and/or a.f. cleaning with a mean of D = 48°, I = 37° (α₉₅ = 5.3°); medium and high coercivity remanence resides in large magnetite grains and fine, predominantly hematite, rods in feldspar megacrysts. The Nordingrårapakivi granite yields a mean, also including dual polarities, of D = 221°, I = ?25° (α₉₅ = 13°), and the Gävle granite yields a mean of D = 26°, I = 17° (α₉₅ = 13°).New data define the a.p.w. path for the Baltic Shield after final uplift and cooling of the ca. 1800 m.y. Svecofennian mobile belt and prior to intrusion of the post-Jotnian dolerites at 1250 m.y.; this (ca. 1500–1200 m.y.) path defines a double loop similar in size and shape to the contemporaneous path for the Laurentian Shield and the paths can be superimposed to define relative positions of the shields. They were in juxtaposition prior to 1200 m.y. with the optimum reconstruction obtained by rotation of approximately 64° about a Euler pole at 1°E, 36°N. Pre-1500 m.y. palaeomagnetic data are also shown to fit this same unique reconstruction. The main geological correlations are an alignment of the Lower/Middle Proterozoic major strike-slip zones, the structural trends within the pre-1700 m.y. mobile belts, and the Grenville and Sveconorwegian (ca. 1100 m.y.) mobile belts. The anorogenic magmatism characteristic of Proterozoic times became gradually more restricted to one active margin of the continental reconstruction as temperature gradients decreased and the crust consolidated. All of these Proterozoic tectonic/magmatic trends are parallel to the long axis of the continental reconstruction.     

Sulitjelma Gabbro,northern Norway: A palaeomagnetic result

The Sulitjelma Gabbro situated at 67.2°N, 15.4°E was intruded close to the Late Ordovician climax of regional metamorphism in the northern Scandinavian Caledonides. Magnetic properties have been examined from samples collected at seven localities in the south western part of this body. Total NRM directions show a tendency to be aligned near the present earth's magnetic field direction in this region. Stability to a.f. demagnetisation is low and commensurate with low Koenigsberger ratios (0.001–0.16) and the presence of unoxidised magnetite as the principal remanence carrier. After cleaning the site mean directions no longer show an alignment near the present earth's field and of six statistically significant sites three are approximately reversed with respect to remainder. The combined mean direction after cleaning isD = 195°,I = 15° (precision parameterk = 6) and the derived virtual geomagnetic pole is at 0°E, 14°S (α₉₅ = 23°). This pole is close to poles of comparable age from the British Isles and suggests that Britain and Norway were part of the same crustal plate in Ordovician times. Discrepancies between Siluro-Devonian results from the two regions may be due to inadequate age coverage of present results.     

A new palaeomagnetic result from East Africa and estimates of the Mesozoic palaeoradius

Of 16 sites collected in the Taru grits (Permian) and Maji ya Chumvi beds (Permo-Triassic) of East Africa only 6 sites from the Maji ya Chumvi sediments gave meaningful palaeomagnetic results. After thermal cleaning the 6 sites (32 samples) give an Early Triassic pole at 67°N, 269°E with A₉₅ = 17° in excellent agreement with other African Mesozoic poles. There are now 26 Mesozoic palaeomagnetic poles for Africa from widely diverse localities ranging in present latitude from 35°N to 30°S. The poles subdivide into Triassic (17 poles) and Cretaceous (9 poles) groups whose means are not significantly different. The palaeomagnetic pole for Africa thus remained in much the same position for 170 m.y. from Early Triassic to Late Cretaceous. The data form an especially good set for estimating the palaeoradius using Ward's method. Values of 1.08 ± 0.15 and 1.03 ± 0.19 times the present radius are deduced for the Triassic and Cretaceous respectively with a mean value of 1.08 ± 0.13 for all the Mesozoic data combined. The analysis demonstrates that hypotheses of earth expansion are very unattractive.     

Palaeomagnetism of silurian lavas of Somerset and Gloucestershire,England

Detailed alternating field demagnetisation of Upper Llandovery volcanics of the Mendip Hills and Gloucestershire has isolated remanence directions interpreted as primary from each of five sites. Well-defined high-coercivity secondary magnetisation is present in six samples of one site and low-coercivity secondary remanence is present in all samples from another site; the former component was apparently acquired in Permo-Triassic times. Primary directions of magnetisation show marked improvement in precision after correction for penecontemporaneous folding, and show a late Llandovery reversal in the sense R → N.The group mean directions of magnetisation isD = 243.5°,I = 47.5° (precision parameterk = 29). Petrographic examination confirms observations from magnetic properties that relict titanomagnetite (oxidation classes 3 to 5) is the remanence carrier in most samples. Hematite, probably mostly late magmatic in origin, is widely developed in all samples, but only the principal remanence carrier where it has thoroughly replaced the titanomagnetite. Low-coercivity remanence is apparently caused by weathering effects but there is no clear visible cause for secondary high-coercivity remanence carried by some samples.The mean virtual geomagnetic pole position is close to Upper Silurian/Lower Devonian pole positions from other parts of Britain and defines a minimum apparent polar shift of 60° between late Ordovician and Upper Llandovery times. Reference to absolute age dates suggests that this shift took place between ca. 447 and 434 m.y. followed by slight polar movement between ca. 434 and 394 m.y.     

Paleomagnetism of Lower Devonian volcanics and Devonian dykes from northcentral New Brunswick,Canada

Some 50 oriented samples (120 specimens) have been collected on eight sites of volcanic rocks from the Lower Devonian Dalhousie Group of northern New Brunswick and Devonian andesitic to basic dykes from central New Brunswick. Univectorial and occasional multivectorial components were extracted from the various samples. Results after AF and thermal demagnetization compare relatively well. In the volcanics and tuffs, two components of magnetization have been isolated: A (D = 33°, I = ?58°, α₉₅ = 7.3°, K = 236) for four sites and B (D = 66°, I = +53°) for three sites. The grouping of component A is improved after tilt correction but the fold test is not significantly positive at the 95% confidence level. Component A is interpreted as being primary while component B is unresolved and appears to be the resultant magnetization of a Late Paleozoic and a recent component. The pole position obtained for tilt corrected component A is 268°E, 1°S, d_p = 6.5°, d_m = 8.8°. The paleolatitude calculated for component A is 39°S. The paleopole of in situ component A is located close to those of the Early-Middle Devonian formations from Quebec, New Brunswick and New England states while the paleopole of tilt-corrected component A is similar to Lower Devonian poles of rock units from the Canadian Arctic Archipelago. If component A is primary (as we believe it to be), then the western half of the northern Appalachians had already docked onto the North American Craton by Early Devonian time. Alternatively, if component A is secondary the same conclusion applies but the juxtaposition took place in Middle Devonian time.     

Palaeomagnetism of Mesozoic igneous rocks from the Maranha˜o Basin,Brazil, and the time of opening of the South Atlantic

From Middle-Upper Jurassic volcanics at the western margin of the Maranha?o Basin (6.4°S, 47.4°W) 15 sites (121 samples) have a mean magnetization directionD = 3.9°,I = ?17.9° withα₉₅ = 9.3°,k = 17.9 after AF cleaning (all sites have normal polarity). This yields a pole (named SAJ₂) at 85.3°N, 82.5°E (A₉₅ = 6.9°) which is near to the other known Middle Jurassic South American pole. For 21 sites (190 samples) from Lower Cretaceous basalt intrusions from the eastern part of the Maranha?o Basin (6.5°S, 42°W) the mean direction isD = 174.7°,I = +6.0° withα₉₅ = 2.8°,k = 122 (all sites have reversed polarity) yielding a pole (SAK₉) at 83.6°N, 261°E (A₉₅ = 1.9°) in agreement with other Lower Cretaceous pole positions for South America. Comparing Mesozoic pole positions for South America and Africa in the pre-drift configuration after Bullard et al. [13] one finds a significant difference (with more than 95% probability) for the Lower Cretaceous and Middle Jurassic poles and also a probable difference for the mean Triassic poles indicating a small but probably stationary separation of the two continents from the predrift position in the Mesozoic until Lower Cretaceous time which may be due to an early rifting event.     

Palaeomagnetism of Cretaceous and Jurassic volcanic rocks in west Sicily

The mean palaeomagnetic pole position obtained from Upper Cretaceous rocks in west Sicily is at 21°N, 100°E (A₉₅ = 15°), and at 38°N, 67°E (A₉₅ = 31°) obtained from Middle Jurassic rocks. These pole positions are completely different from comparable pole positions for southeast Sicily and Africa and imply a clockwise rotation of west Sicily since the Upper Cretaceous of about 90° relative to southeast Sicily and Africa and also a clockwise rotation of about 60° relative to “stable” Europe. The sense of rotation of west Sicily is opposite to any known rotation of other crustal blocks in the central Mediterranean.     

The palaeoposition of Madagascar: Palaeomagnetic evidence from the Isalo group

Palaeomagnetic results are reported from the continental facies of the Triassic-Jurassic Isalo Group of Madagascar. Stability of the magnetic remanence was tested using the alternating field and progressive thermal demagnetization techniques. Results from 8 sites, 6 located in northwestern Madagascar and 2 from the southwestern region, yield a palaeomagnetic pole at 74.2°S, 97.1°E (N = 8, A₉₅ = 6.3°). Three models previously proposed to describe the drift history of Madagascar relative to Africa are considered and the relevant geological and geophysical information is reviewed. The palaeomagnetic data are only consistent with the pre-drift model which places Madagascar off the east coast of Africa adjacent to Kenya and Tanzania. This is also the continental drift fit favoured on geological grounds.