V. V. Popov and A. N. Khramov
Department of Palaeomagnetic reconstructions, All-Russia Petroleum Research and Geological Exploration Institute (VNIGRI), St. Petersburg, Russia
Department for Earth and Environmental Sciences, Geophysical Section, Ludwig-Maximilians University, Munich, Germany
The Late Neoproterozoic palaeogeography attracted the attention of many researchers in connection with the hypothesis of the Rhodinia supercontinet which had begun to break up about 725 million years ago. The views on the subsequent events, such as, the unusually rapid drifting of the continents, their low-latitude positions during the glaciation epochs, and their connection to form the Palaeozoic Pangea, were based mainly on palaeomagnetic data which were extremely scarce for all of these continents. For instance, some authors [Torsvik et al., 1996] used merely five palaeomagnetic poles for the Baltic Vendian rocks, which satisfied the modern criteria of reliability [Van der Voo, 1990]. It is clear that these data are absolutely insufficient for choosing any of the competing models available for the movements of the Baltic continent which occupies the central key position in the modern global reconstructions for the Vendian time. As follows from the data reported in [Torsvik et al., 1996], the southern palaeomagnetic pole of the Baltic region had been located in the Vendian time at high northern latitudes. These data suggest that during the Late Vendian time the Baltic region had resided in the region of 60oS. The new palaeomagnetic datings obtained recently for the Vendian rocks of the Russian Platform [Popov, 2001; Popov and Khramov, 2002; Popov et al., 2002; Iglesias et al., 2004], as well as the results of this study, suggest a new version for the palaeogeographic position of the Baltic Continent during Vendian time.
The Vendian rocks of the region outcrop along the eastern margin of the Baltic Schield in the river valleys of the Onega Range, in the western segment of the Dvina scarp, and in the west of the White-Sea Kuloi scarp, as well as in the scarps of the Onega, Letnii, and Zimnii shores of the White Sea. The Vendian rocks rest almost horizontally on the crystalline basement of the platform and the Upper Riphean rocks. The Late Vendian rocks of the southeastern White Sea region are known to show the best rock sequence in the Russian Platform [Stankovski et al., 1985]. The Ediacarian fossils, unique in their number, variety, and preservation [Fedonkin, 1981, 1985; Grazhdankin, 2000; Grazhdankin and Bronnikov, 1997], as well as the absolute age determinations using zircons [Iglesias et al., 2004; Martin et al., 2000; Shchukin et al., 2002], date very precisely the various parts of the rock sequence, the total thickness of which is greater than 500 m.
The samples were collected from 10 outcrops, partially overlapping one another, in the middle course of the Zolotitsa River (from 65o20 N and 40o00 E to 65o40 N and 40o30 E, Figure 1) and from the vertically oriented cores collected from a borehole. The total study interval included 90 meters of the Zolotitsa River section with a sampling interval ranging from 5 cm to 100 cm (350 oriented lump samples) and 420 m of holes cores collected with a sampling interval of 40 cm to 100 cm (443 samples). Where the rocks were studied in the outcrops, they showed subhorizontal bedding with dip angles smaller than 10o. Only the lower sequence of rocks exposed by the Zolotitsa River showed two anticlinal folds with the maximum dip angles of 70o. The folds are believed to have been produced by recent glacial dislocations. The oriented lump samples collected from the Zolotitsa River site were sawed into cubes with the edges of 2 cm (880 cubes); the core samples were used to collect cubic samples with a side of 2 cm or cylinders, 22 mm long and 25 mm in diameter, collecting one sample from each stratigraphic level.
The palaeomagnetic and petromagnetic studies of the samples were carried out in the VNIGRI Department for Palaeomagnetic Reconstructions (St. Petersburg) and in the Department for Earth and Environmental sciences, Geophysical Section, Ludwig-Maximilian University (Munich, Germany). The NRM of the samples was measured using JR-4 and JR-5 spinner magnetometers (St. Petersburg) and a 2G cryogenic magnetometer (Munich). The samples were subject to detailed cleaning using an alternating magnetic field and to thermal cleaning. The cleaning by the alternating magnetic field did not give any satisfactory results; most of the samples experienced thermal demagnetization. Cleaning was performed in Saint Petersburg using stoves of the VNIGRI construction and TD-48 stoves (USA) and in Munich (Germany) using TSD-1 stoves (Schonstedt production) and TD-48 (USA). The results of magnetic cleaning were analyzed using Ziyderveld diagrams [Zijderveld, 1967] and stereoprojections. The linear segments corresponding to NRM components were located on the demagnetization curve in agreement with the technique of a principal component analysis [Kirschvink, 1980]. The average directions of the resulting magnetization components were computed using the Fisher Statistics [Fisher, 1953]. Where several samples, sawed from the same bulk sample, were used, the directions of their components were averaged, that is, all statistical data reported in this paper are given at the level of lump samples. The computer processing of the results was accomplished using the programs [Enkin, 1994]. To control the chemical variations of NRM carrying minerals the magnetic susceptibility of the samples was measured using a KLY-2 Kappa bridge. The cleaning of the samples was terminated, where their magnetic susceptibility increased 5 or more times, or where the value and direction of the residual vector showed chaotic changes at several cleaning steps. These changes are typical of gray rocks and were observed occasionally in red sandstones.
The compositions of magnetic minerals (NRM carriers) were identified in Munich using a thermomagnetic analysis and hysteresis characteristics obtained using a VFTB (Variable Field Translation Balance) equipment, and also in Saint Petersburg using the curves recording changes in magnetic susceptibility as a function of temperature, recorded by a KLY-3 Kappa bridge and a CS-3 thermal attachment (Spinning Specimen Magnetic Susceptibility Anisotropy Meter). This instrument was also used to study the anisotropy of the magnetic susceptibility of the samples collected from the Zolotitsa River cross section.
The results of the thermomagnetic analysis of the samples collected from the Zolotitsa River rock sequence showed that they consisted of hematite alone or the latter was obviously a predominant mineral. The content of magnetite (or, to be more exact, of oxidized magnetite, because none of the samples showed Tc = 580oC) was higher in the sandstones of the lower (Vendian) rock sequence. In most cases the magnetization of the samples grew notably in the course of their heating, beginning from the temperature of about 200oC and showing the maximum values at ~400oC. The J(T) peaks in the upper part of the rock sequence were recorded in all samples, except for those from the Vendian interval where the peaks diminish with some samples showing a reversed J(T) curve pattern. A similar effect is often observed in experimental J(T) curves, measured in the absence of any saturation field, especially in the case of hematite-bearing rocks [e.g., Duff, 1979]. The theoretical substantiation of the peaks in the thermal magnetization and susceptibility curves observed for a great number of compounds was offered by I. Y. Korenblit and E. F. Shender , and this phenomenon was referred to as "spin glass''. Later, the model of the spin glass behavior was applied to a system of interacting single-domain particles disseminated in a nonmagnetic matrix resembling a natural rock material [Belokon and Nefedov, 2001; Duff, 1979]. The presence of a peak in the J(T) curve can be explained in the following way. In the case of disorded hematite ("spin glass''), and in the case where the field in which the thermomagnetic analysis was performed was not a saturation field (the thermomagnetic analysis was performed in the field of 100 mT, the saturation field being higher than 250 mT), the magnetic material was ordered during the heating, following the ordering of the spin glass type, leading to the growth of magnetization which was preserved during the cooling in the field, producing an irreversible J(T) curve. As to the samples with the notable contents of magnetite, oxidized magnetite, and some magnetically "soft'' material, the value of 100 mT denotes a saturation field and suggests these minerals to be magnetically ordered and showing no effect concerned. Moreover, since these minerals are more magnetic than hematite, the effect of hematite is masked. The ubiquitous presence of J(T) peaks suggests the predominance of more rigid hematite with a disorded magnetic structure in the upper part of the rock sequence. All of the studied samples showed their second heating and cooling J(T) curves, almost coinciding with the first cooling curve, confirming the explanation of the J(T) behavior offered in this paper.
The k(T) variation curves also prove the presence of hematite in the study rocks. The curves obtained for the first and second heating showed substantial differences in many cases (Figure 8). The data obtained for thermal NRM demagnetization confirm the petromagnetic data, namely, the fact that the carrier of its high-temperature component in the rocks is hematite alone.
The predominance of hematite in the rocks of the whole sequence is proved by the high magnetic rigidity of the rock samples: magnetic saturation is not achieved in the field of up to 250 mT, Hcr = 110-170 mT. As to the Vendian rocks, only some samples of gray sandstone showed Hcr<50 mT (e.g., only one sample (no. 200) showed Hcr = 30 mT. Its J(T) curve showed Tc = 610oC and 680oC, suggesting the presence of single-phase oxidized magnetite and hematite). There are also indications of maghemite (a J peak in the vicinity of 200oC in red claystone sample no. 343). The sandstone samples collected from the top of the rock sequence, showing the lowest k values and the highest Q values, are distinguished by a broad, flattening hysteresis loop, the irreversible J growth during the heating, a significant difference between the J values before and after the heating, and a segment with negative J values in the J(T) curve (Figure 9a). The claystone and siltstone samples collected from the lower part of the rock sequence, distinguished by the lowest k and highest Q values, showed a more narrow hysteresis loop, the absence of any segments with negative J values in the J(T) curve, a lower J(T) peak, and the lower difference between the J values before and after the heating (Figure 9b). These data suggest that this interval of the rock sequence is composed mostly of coarse-grained hematite which is accompanied by some magnetite. The presence of oxidized magnetite and maghemite in the Vendian rocks, even in small amounts, can explain the low Q values in these rocks, compared to the Cambrian "purely hematitic'' sediments, apart from the role of diamagnetic carbonate rocks.
The A component was found to disintegrate in the narrow temperature range of 20-250o C (Figure 10) and in many cases in the temperature range of 20-150oC (Figure 10) and was identified in all rock samples. This component showed the best statistics where a progressive fold test was used (with 55% rectification) [McFadden and Jones, 1981; Watson and Enkin, 1993] and yielded a mean direction which almost coincided with the direction of the present-day geomagnetic field at the sampling site (Figure 11a, and Table 1). It appears that this component is the product of modern (partially, laboratory) remagnetization.
The mean-temperature B component showed blocking temperatures, varying from sample to sample. In some cases it was destroyed at a temperature range of 100-300oC, in other cases it remained to be stable at the temperatures as high as 680oC (Figure 10). This component was recorded in all rock samples and showed the best statistical values, both where a progressive fold test was used, and where the whole collection of samples was used (40% rectification) [McFadden and Jones, 1981; Watson and Enkin, 1993], and where we only used the samples where the rocks had substantially different dips and strikes (50% rectification). The palaeomagnetic pole obtained in the latter case had the following coordinates: F = 25.8oN, L = 44.2oE, dp = 2.5o, and dm = 3.0o (Figure 11b, Table 1).
The medium-temperature C component was identified in a small number of samples, namely, in 39 samples out of the studied 350 lump samples. All of these samples had been collected in the lower segment of the Zolotitsa River section, where the rocks had been folded, and also below this zone. This component was located in a temperature range of 400-670oC and always showed higher blocking temperatures, compared to the those of the B component (Figure 10). The C component also showed the best distribution in the stratigraphic system of the coordinates and a positive fold test with a probability of 98% [McFadden and Jones, 1981; Watson and Enkin, 1993] (Figure 11c, and Table 1). The palaeomagnetic pole ( F = 38.5oS, L = 356.2oE, dp = 3.7o, and dm = 6.5o) corresponding to this component is located in the vicinity of the Permian pole for the Russian Platform [Iosifidi and Khramov, 2002; Smethurst et al., 1998; Torsvik et al., 1996].
The Z high-temperature component was identified in many of the study samples (234 lump samples). It was destroyed in the narrow temperature range of 660-690oC and occasionally in the range of 680-695oC (Figure 10). In most cases the recording of this component called for diminishing the interval of the temperature cleaning at the last steps down to 5oC, 3oC, and 2oC, up to 20-50% of NRM being often destroyed at the last ten degrees of cleaning (Figure 10). The Z component distinguished in this way showed two polarities. Its polarity was not controlled by the lithology of the samples (being most reliably detected in the red and mottled clays) or by the component composition of a particular sample (in the presence of all components or of merely A, B, or Z), but was controlled only by the positions of the samples in the rock sequence. The Z component showed a positive reversal test of the B class (its difference from the antiparallel pattern of the direct and reverse mean directions being 4.1o with the critical angle of 6.7o ) [McFadden and McElhinny, 1990]. The fold test was also found to be positive with a probability of 99% [McFadden and Jones, 1981; Watson and Enkin, 1993] for the components of direct polarity, reversed polarity, and for all population of the vectors (Figure 11d, and Table 1). The palaeomagnetic pole ( F = 31.7oS, L = 112.9oE, dp = 1.6o, dm = 2.7o), calculated from the mean direction of the Z component resides far from the trajectory of the apparent migration of the palaeomagnetic pole of the Russian Platform [Smethurst et al., 1998; Torsvik et al., 1996] and is close to the pole which was obtained earlier using the Zimnegorsk outcrops at the Winter Coast of the White Sea [Popov et al., 2002]. All of the above data allow us to rank the Z component as the old and primary component, synchronous to the sediments accumulation.
We could not use the low-temperature modern component of magnetization to orient the cores in space using the conventional method for several reasons.
In the case of this situation we offer a new method for orienting drill cores in space, which allows one to calculate, adhering to certain conditions, the direction of the modern magnetization component in the cases where the conventional method fails to give a satisfactory result for the reasons mentioned above. The basic condition for using this new method is the availability of the results of the magnetic cleaning of the rocks of the same type and age, collected from natural outcrops and drill cores, and the proved identity of the NRM component composition of these rocks. In this case the cores should be oriented using the directions of the lines normal to the circles drawn via the directions of the old components, rather than the modern viscous component. The procedure of this work is as follows:
(1) chose two most representative magnetization components, found in the predominant number of the samples collected from natural outcrops: A1 and A2;
(2) compute the mean directions of these components: DA1, IA1 and DA2, IA2;
(3) compute the mean direction of the normal to the large circle drawn via these average Dn and In directions;
(4) compute the direction of a normal to the large Dni, Ini circle drawn via the similar A1i and A2i components;
(5) compute the inclination difference Di = Dn - Dni between the average direction of the Dn, In normal and the direction of the normal for each i th core sample Dni, Ini;
(6) add the computed Di value to the inclinations of all components which have been identified in the ith sample.
This procedure allows one to compare the reconstructed distributions in the drill cores and in the natural outcrops for several magnetization components simultaneously.
The correctness of this method is based on the following statement: if the application of this method for the distributions obtained in natural outcrops does not modify the average directions of the magnetization components within an error, this method can be used for similar components identified in core samples.
1. the computation of the D n, J n normal to the large circle drawn via the average directions of the B and Z components (Table 1);
2. the random selection of 55 samples showing the direct polarity of the Z component (N group) and 55 samples showing the inverse polarity (R group), Figure 13;
3. the chosen vector samples were processed using the above technique and the B and Z components, Figure 14;
4. the average directions and statistical parameters of the B and Z components, as well as the angle d between these average directions, were computed for the N and R groups prior to and after using this method;
5. 5, 10, 50 random samples were discarded from the N and R groups, and the computations mentioned in item 4 were repeated (Figure 15);
As can be seen in Figure 15, the differences between the average d directions prior to and after the use of this method for all data samples, up to the diminishing the number of samples to 5 in each group, are not greater than the angle of confidence a95 for each data distribution, both for the B and Z components. This proves the correctness of using this new method.
More than 800 samples, collected from natural outcrops and from drill hole cores were subjected to a complete cycle of palaeomagnetic studies. More than 400 m of the classical Late Vendian-Early Cambrian rock sequences were studied in the north of the Russian Platform. The rocks were dated using the Ediacarian fauna found in the Zimnegorsk outcrops and also using zircons found in the lower part of the Zolotitsa River section (550 Ma). The results of our study proved the multicomponent NRM composition. Two medium-temperature components, showing different blocking temperatures, were proved to have been associated with the different-age remagnetization of the rocks. The high-temperature bipolar component showed positive fold, reversal, and correlation tests. Its carrier was found to be hematite. 27 zones of the direct and 28 zones of the reversed polarity were located in the Kotlin Horizon of the Late Vendian-Early Cambrian rocks using the direction of this component. A new technique was offered and tested for orienting borehole cores in space. Its use allowed us to reconstruct the distribution of NRM components in the core samples. The resulting distributions coincided with the distributions obtained for natural outcrops. The high reliability of the resulting data makes it possible to introduce substantial corrections to the potential positions of the apparent migration path of the Baltic palaeomagnetic pole for Vendian-Cambrian time and to reconstruct the palaeogeographic positions of the Baltic and Laurentia for the investigated period of geological time.
Belokon, V. I., and K. V. Nefedov (2001), The distribution of random interaction fields in disorded magnetic materials: Spin and macrospin glass, J. Experimental and Theoretical Physics, 120b(1(7)), 156-163.
Bylund, G. (1994), Palaeomagnetism of the Vendian-Early Cambrian sedimentary rocks from E. Finmark, Norway, Tectonophysics, 231, 45-57.
Claesson, K. C. (1978), Swedish Ordovician limestones: Problems in clarifying their directions of magnetization, Phys. Earth Planet. Inter., 16, 65-72.
Duff, B. A. (1979), Peaked thermomagnetic curves for hematite-bearing rocks and concentrates, Phys. Earth Planet. Inter., 19, 1-4.
Enkin, R. J. (1994), A Computer Program Package for Analysis and Presentation of Palaeomagnetic Data, 16 pp., Pacific Geoscience Center, Geol. Survey of Canada, Sidney.
Fedonkin, M. A. (1981), White Sea Vendian Biota (in Russian), 98 pp., Nauka, Moscow.
Fedonkin, M. A. (1985), Systematic description of Vendian metazoans, in Vendian System, Geological History and Palaeontology, vol. 1. Palaeontology (in Russian), pp. 70-106, Nauka, Moscow.
Fisher, R. A. (1953), Dispersion on a sphere, Proc. R. Soc, London, Ser. A, (217), 295-305.
Grazhdankin, D. V., and A. D. Bronnikov (1997), New finds of late Vendian fossil remains in the Onega Peninsula, Dokl. Akad. Nauk, 357(6), 792-796.
Grazhdankin, D. V. (2000), The Ediacaran genus Inaria: A taphonomic/morphodynamic analysis, Neues Jahrb. Geol. Palaeontol Abh., 47, 279-308.
Iglesias, M. P. L., J. A. Tait, V. V. Popov, and A. Abalmasova (2004), Palaeomagnetic data from the Vendian sediments of the Arkhangelsk region, NW Russia: An alternative apparent polar wander path of Baltica for the Late Proterozoic-Early Palaeozoic (submitted).
Iosifidi, A. G., and A. N. Khramov (2002), Palaeomagnetism of the Upper Carboniferous and Lower Permian deposits of the East European Platform: A key palaeomagnetic pole and kinematics of collision with the Urals, Phys. Solid Earth, 38(5), 389-403, translated from the Russian Journal Physics of the Earth, (4), 42-56, 2002.
Iosifidi, A. G, A. N. Khramov, and V. Bachtadse (2005), Multicomponent magnetization of the Vendian sedimentary sequences of Podolia, Ukraine, Russian J. Earth Sci., 7(1), 1-14. http://rjes.wdcb.ru/v07/tje04166/tje04166.htm
Kirschvink, J. L. (1980), The least-squares line and plane and the analysis of palaeomagnetic data, Geophys. J. R. Astron. Soc., 62, 699-718.
Lowrie, W. (1990), Identification of ferromagnetic minerals in a rock by coercitivity and unblocking temperature properties, Geophys. Res. Lett, 17, 159-162.
Martin, M. W., D. V. Grazhdankin, S. A. Bowring, D. A. Evans, M. A. Fedonkin, and Y. L. Kirskhwink (2000), Age of Neoproterozoic bilaterian body and trace fossils, White Sea, Russia: Implications for Metazoan evolution, Science, 288(5), 841-845.
McElhinny, M. W., and J. Lock (1996), IAGA palaeomagnetic databases with access, Surv. Geophys., 17, 575-591.
McFadden, P. L., and D. L. Jones (1981), The fold tests in palaeomagnetism, Geophys. J. R. Astron. Soc., 67, 53-58.
McFadden, P. L., and M. W. McElhinny (1990), Classification of reversal tests in palaeomagnetism, Geophys. J. Int., 103, 725-729.
Meert, J. A., T. H. Torsvik, E. A. Eide, and S. Dahlgren (1998), Tectonic significance of the Fen Province, S. Norway: Constraints from geochronology and palaeomagnetism, J. Geol., 106, 553-564.
Meert, J. A., and R. Van der Voo (1994), Palaeomagnetism of the Catoctin volcanic province: A new Vendian-Cambrian apparent pole wander path for North America, J. Geophys. Res., 106(B3), 4625-4641.
Opdyke, N.D., and J. E. T. Channell (1994), Magnetic Stratigraphy, 45 pp., Academic Press.
Perroud, H., M. Robardet, and D. L. Bruton (1992), Palaeomagnetic constraints upon the palaeogeographic position of the Baltic Shield in the Ordovician, Tectonophysics, 201, 97-120.
Piper, J. D. A. (1981), Magnetic properties of the Alnon Complex, Geol. Foeren. Stockholm. Foerh., 103, 9-15.
Piper, J. D. A (1988), Palaeomagnetism of Late Vendian-Early Cambrian minor alkaline intrusions, Fen Complex, South Norway, Earth Planet. Sci. Lett., 90, 422-430.
Popov, V. V. (2001), Palaeomagnetism of the Upper Vendian sedimentary rocks outcropping along the Zolotitsa River (Winter Coast of the White Sea), in Proc, Seminar: Palaeomagnetism and Magnetism of Rocks, held in Borok in October 18-23, 2001 (in Russian), pp. 71-73, GEOS Press, Moscow.
Popov, V., A. Iosifidy, A. Khramov, J. Tait, and V. Bachtadse (2002), Palaeomagnetism of Upper Vendian sediments from the Winter Coast, White Sea region, Russia: Implications for the palaeogeography of the Baltica during Neoproterozoic time, J. Geophys. Res., 107(11), 2315, doi:10.1029/2001JB001607.
Popov, V., and A. Khramov (2002), Palaeomagnetism of Late Vendian sedimentary rocks from the Winter Coast of the White Sea, based on the study of core samples: A new method for drill core orientation, in: Palaeomagnetism and Magnetism of Rocks, Proc. Seminar, held in Borok in October, 19-22, 2002, GEOS Press. pp.74-76, 2002.
Poorter, R. P. E (1972), Preliminary palaeomagnetic results from the Fen carbonatite complex, S. Norway, Earth Planet. Sci. Lett., 17, 194-198.
Roberts, D., I. M. Gorokhov, A. Siedlecka, N. N. Melnikov, T. I. Turchenko, G. V. Konstantinova, E. P. Kutyavin, and A. V. Sochava (1997), Rb-Sr dating of ielite fractions from Neoproterozoic shales on Varanger Peninsula, North Norway, Norges Geologiske Undersokelse Bulletin, 433, 24-25.
Shchukin, V. S., S. M. Sablukov, L. I. Sablukova, E. A. Belousova, and W. L. Griffin (2002), Late Vendian aerial alkaline volcanism of rifting type in the Zimnii Bereg kimberlite area (Arkhangelsk diamondiferous province), in Deep-Seated Magmatism, Magma Sources, and the Problem of Plumes, pp. 203-212.
Shipunov, S. V., and N. M. Chumakov (1991), Palaeomagnetism of the Late Proterozoic deposits in the Kola Peninsula, Geotectonics, 25(5), 401-410.
Smethurst, M. A., A. N. Khramov, and S. A. Pisarevsky (1998), Palaeomagnetism of the Lower Ordovician orthoceras limestone: A revised drift history for Baltica in the Early Palaeozoic, Geophys. J. Int., 133, 44-56.
Stankovski, A. F., E. M. Verichev, and I. P. Dobeiko (1985), The Vendian rocks of the Eastern White Sea region, in The Vendian System: Historic Geology and Palaeontology, Stratigraphy and Geologic Processes (in Russian), vol. 2, pp. 67-76, Nauka, Moscow.
Tarling, D. H., and F. Hrouda (1993), The Magnetic Anisotropy of Rocks, St. Edmundsbury Press, Suffolk.
Torsvik, T. H., and E. F. Rehnstrom (2001), Cambrian palaeomagnetic data for Baltica: Implications for true polar wander and Cambrian palaeogeography, J. Geol. Soc. London, 158, 321-329.
Torsvik, T. H., D. Roberts, and A. Siedlecka (1995), Palaeomagnetic data from sedimentary rocks and dolerite dikes, Kildin, Rybachii, and Srednii Islands and NE Norway: A review, Spec. Publ., 7, 315-326.
Torsvik, T. H., and M. A. Smethurst (1999), Plate tectonics modeling: Virtual reality with GMAP, Comput. Geosci., 25, 395-402.
Torsvik, T. H., M. A. Smethurst, J. G. Meert, et al. (1996), Continental break-up and collision in the Neoproterozoic and Palaeozoic: A tale of Baltica and Laurentia, Earth Sci. Rev., 40, 229-258.
Torsvik, T. H., and A. Trench (1991a), Ordovician Magnetostratigraphy: Llanvirn-Caradoc limestones of the Baltica platform, Geophys. J. Int., 107, 171-184.
Torsvik, T. H., and A. Trench (1991b), The Lower-Middle Ordovician palaeofield of Scandinavia: Southern Sweden revisited, Phys. Earth Planet. Inter., 65, 283-291.
Torsvik, T. H., R. Van der Voo, J. G. Meert, et al. (2001), Reconstructions of the continents around the North Atlantic at about the 60th parallel, Earth Planet. Sci. Lett., 187, 55-69.
Van der Voo, R. (1990), The reliability of palaeomagnetic data, Tectonophysics, 184(1), 1-9.
Watson, G. S., and R. J. Enkin (1993), The fold test in palaeomagnetism as a parameter estimation problem, Geophys. Res. Lett., 20, 2135-2138.
Zijderveld, J. D. A. (1967), Demagnetization of rocks: Analysis of the results, in Methods in Palaeomagnetism, edited by D. W. Collinson, pp. 254-286, Amsterdam, Elsevier.