Palaeomagnetic and rock magnetic study of Devonian olistoliths set in early Carboniferous flysch, West Sudetes, Poland

Summary

The Upper Devonian greywacke, mudstone and limestone olistoliths from the Zdanów Bardo Młynów (ZBM) allochthonous complex of the Góry Bardzkie (GB), West Sudetes, have been sampled for palaeomagnetic investigations (46 hand samples and 30 drill cores from seven sites). The random distribution of the olistolithic blocks within early Carboniferous flysch provides an opportunity for the megaconglomerate test. Microscopic and rock magnetic studies showed that pyrrhotite and magnetite are the carriers of magnetic remanence. Thermal and AF demagnetization revealed four components of characteristic remanence. The C component, found in one site only, yields an in situ pole position of 19°S, 340°E, close to early Carboniferous data for Baltica. The P component, also found in one site, yields an in situ pole position of 42°S, 339°E, which coincides with late Permian poles for Baltica. The R component of moderate inclinations and northwesterly and northeasterly declinations, found in all but one site, is probably partial remagnetization of a Devonian component during Carboniferous and Permian times. The D component found in all sites shows significant dispersion of declination of site mean directions in geographical and stratigraphic coordinates, but inclinations after tilt correction are consistent with inclinations expected for the Middle–Late Devonian calculated for the Sudetes from the palaeopoles for Baltica. Significant dispersion of declination is expected for primary magnetization of olistoliths distributed at random within flysch. The fold test for inclination only, applied to the site mean directions of the D component for the ZBM complex, is positive. The D component is therefore interpreted as representative of the Middle–Late Devonian local field inclination. The calculated palaeolatitudes, 15°S–7°S, indicate that the West Sudetes were positioned in that time at the same latitude as the southwestern margin of Baltica.

Introduction

The West Sudetes form the northeastern part of the Bohemian Massif belonging to the central European Variscides (Fig. 1). The Variscan orogenic belt of Europe is a collage of Gondwana derived terranes. These terranes formed two semi rigid plates: Avalonia and Armorica, they themselves being a mosaic of several microplates and blocks referred to as the Armorican Terrane Assemblage (ATA; Tait et al. 1997). Avalonia drifted to the north and collided with Baltica in late Ordovician–early Silurian time (Torsvik et al. 1992, 1993; Tait et al. 1994, 1997; McKerrow & Cocks 1995). Having been separated from Avalonia by the Rheic Ocean, the ATA drifted independently northwards. Palaeomagnetic and palaeobiogeographic data indicate that the Rheic Ocean was closed in late Silurian–early Devonian time (Cocks et al. 1997), when the Armorican terranes joined Avalonia.

The Rheic suture zone between Avalonia and Armorica is mostly concealed by younger sediments, hence its position is not clear. It is placed roughly along the northwestern edge of the Mid German Crystalline High, representing by itself the northern part of the Saxothuringian plate (Tait et al. 1997). Further east, Oliver et al. (1993) and Cymerman & Piasecki (1994) speculatively positioned this suture inside the West Sudetes and interpreted it as a Caledonian collision zone. Such views are, however, criticized as neglecting the basic stratigraphic and structural features of the Sudetes (Franke et al. 1993; Aleksandrowski 1994; Żelaźniewicz & Franke 1994; Żelaźniewicz 1997). Preliminary palaeomagnetic studies of the lower Ordovician metagranites and metavolcanic rocks occurring to the south and north of the postulated suture are not conclusive about latitudinal separation across this suture during the Ordovician (Nawrocki & Żelaźniewicz 1996).

The area of Rheic suture was rifted again during the early mid Devonian and re opened as the Rhenohercynian basin with the Mid German Crystalline High and the remainder of the Saxothuringian terrane on its southern margin (Franke et al. 1995; Franke & Oncken 1995). At the same time within the ATA the Tepla Barrandian terrane rotated about 140° anticlockwise and upon closing the Saxothuringian ocean collided with the Saxothuringian terrane by the Famennian (Tait et al. 1994, 1997). New palaeomagnetic data from Silurian carbonate rocks from the Saxothuringian basin (Tait et al. 1997) did not demonstrate such post Silurian rotation. This indicates that the Saxothuringian basin must have been a separate microplate. Accordingly, major oceans separating fragments of central Europe became closed in late Devonian times, which gave way to large scale shearing and transpressional movements during the late Devonian–Carboniferous.

Palaeomagnetic data for that period are sparse in Europe. For Baltica, there is only an Early Devonian pole from Ukraine (Smethurst & Khramov 1992) and there are data for Spitsbergen (Jeleńska & Lewandowski 1986). For Armorica and Avalonia only two poles from the Harz Mountains and the Franconian Forest (Bachtadse et al. 1983) represent the primary magnetization for Central Europe. The palaeolatitudes of 13°S obtained from these poles are similar to those expected from Ukrainian and Spitsbergen data. These data indicate the proximity of Armorica to the southwestern margin of Baltica from the early Devonian onwards.

The palaeomagnetic results obtained from the Sudetic ophiolite dated using the U–Pb method on zircon grains at c. 420 Ma (Oliver et al. 1993) or c. 400 Ma (Żelaźniewicz et al. 1998) of the latest Silurian–early Devonian interval revealed, besides a Carboniferous overprint, a direction close to the expected Silurian direction (Jeleńska et al. 1995). Unfortunately, a fold test was not possible, so there is no precise time constraint for this direction. Otherwise the rocks bear Carboniferous magnetic and isotopic (Sm–Nd whole rock of 350 Ma, Pin et al. 1988) overprints.

Therefore, palaeomagnetic studies of pre Carboniferous rocks in the West Sudetes are of great importance as they should help to elucidate the geotectonic history of the Sudetes and Armorica. In this study, we continue our palaeomagnetic investigations in the Góry Bardzkie (GB) region, where olistoliths of various sizes (up to kilometre scale, Haydukiewicz 1998) of mostly upper Devonian rocks occur within Carboniferous flysch (Wajsprych 1995, 1997). As rocks suitable for the fold test are difficult to find in the Sudetes, we expect that the olistoliths, which have experienced random rotations in a turbiditic environment, will allow for the conglomerate test to be applied on the megascale (Tarling 1983; Parés et al. 1994; Tarduno et al. 1990). The conglomerate test can be considered positive if the dispersion in declination is accompanied by consistency in tilt corrected inclinations. Such consistency will be checked by the fold test for inclination only. The positive fold test can provide information about the latitudinal position of the Sudetes in Devonian times. The preliminary results of a palaeomagnetic study for two localities from the ZBM complex have been published by Edel et al. (1997).

Geological setting

Góry Bardzkie is a small tectonostratigraphic unit of the West Sudetes (Fig. 1). It represents a fault bounded fragment of a once larger sedimentary basin. This tectonostratigraphic unit contains practically unmetamorphosed to anchimetamorphosed upper Ordovician(?) and Silurian to lower Carboniferous strata occurring in two major lithotectonic units (Wajsprych 1995). One unit consists of an autochthonous platform succession of late Famennian–Visean age, with clastic components derived directly from the underlying Góry Sowie Block (GSB) and Kłodzko Metamorphic Unit (Fig. 1), passing upwards into flysch sediments. The other unit comprises upper Ordovician–upper Devonian cherty to turbiditic rocks referred to as the Zdanów Bardo Młynów (ZBM) allochthonous complex, which was transformed into a sedimentary accretionary prism, enclosed in the early Carboniferous flysch of the former unit, and thrust over it during the late Visean (Wajsprych 1995, 1997). Accordingly, fragments of this complex occur as olistoliths and slide sheets of different sizes (olistostromes to kilometre scale blocks) within the Carboniferous flysch. The Bardo Młynów (BM) succession, part of the ZBM complex, has been sampled in several places for our palaeomagnetic study (Fig. 1). 46 hand samples and 30 cores were collected from limestones of ?late Devonian age from Podzamek (Fig. 1, site PZ) and the upper Devonian greywacke and greywacke sandstones (Chorowska 1990; Haydukiewicz 1998) from Dębowina (site DE), from a local road near Bardo (site PB) and from a roadcut south of Bardo (site ZB). Middle–upper Devonian greywacke and mudstone were sampled at Młynów (Fig. 1), two from overturned layers (site M2 and M3) and one from a right side up layer (site M1). Four–six specimens were cut from hand samples and one–three specimens were cut from cores.

Mineralogy and rock magnetic study

A study of polished sections under an ore microscope revealed that all the rocks under study contain a variety of pyrite. The pyrite occurs as single automorphic grains inside the matrix or aggregates of framboidal grains. However, pyrrhotite is not observed under the microscope. In some specimens from Młynów (sites M1, M2 and M3) single detrital magnetite and ilmenite grains were recognized. In two other sites, Dębowina (site DE) and the road near Bardo (site PB), pyrite was partly altered to Fe hydroxides, which were seen under the microscope as pseudomorphs after pyrite.

The set of rock magnetic experiments performed to identify magnetic minerals comprised continuous thermal demagnetization of saturation remanence I_rs, step by step thermal demagnetization of isothermal remanent magnetization (IRM) acquired in three perpendicular directions (Lowrie 1990), and IRM and anhysteretic remanent magnetization (ARM) acquisition curves. Rocks are too weak for Curie balance measurements. Hysteresis parameters were measured for a few specimens only because the Molyneaux vibrating magnetometer VSM available in the Warsaw lab for hysteresis loop measurements has insufficient sensitivity for the majority of samples. Continuous thermal demagnetization of I_rs was carried out using a home made device. A specimen was magnetized in a field of 1T. Remanence was then measured during heating in a field free space. The three perpendicular components of remanence used for step by step thermal demagnetization were acquired in a field of 0.11 T along the y axis, 0.4 T along the x axis and 3 T along the z axis. ARM was acquired in a steady field of 0.05 mT and a peak alternating field increasing up to 100 mT.

I_rs(T) curves revealed variable proportions of pyrrhotite and magnetite in greywackes and mudstones from all sites except site PZ (Podzamek limestones). This is evidenced by blocking temperatures of 320 °C and 550–570 °C, respectively (Figs 2a and b). Podzamek (PZ) limestones contain only pyrrhotite (Fig. 2c). After heating in air to a temperature of 700 °C, new magnetite is produced, probably from pyrrhotite and pyrite. Production of magnetite is evidenced by the I_rs(T) curves of the second run. Oxidation of pyrrhotite and pyrite to magnetite generally starts between 350 and 400 °C, with the peak value between 500 and 550 °C. An increase of susceptibility observed in these temperature ranges for PZ limestones suggests magnetite production in such a way. In some samples one phase—magnetite—was observed (Fig. 2d). In spite of microscopic observations of Fe hydroxides, goethite, the magnetic one, was never seen on I_rs(T) curves.

Thermal demagnetization of the three IRM components demonstrates similar behaviour (Fig. 2 insets). For samples containing only magnetite, soft and a small amount of medium components were completely unblocked between 550 and 600 °C, and are clearly due to magnetite (Fig. 2d inset). The small hard component unblocked at 680 °C, indicating haematite as a carrier. For samples from PZ limestones containing only pyrrhotite, the hard component of IRM is negligible. The soft component and the small medium component both unblocked between 320 and 350 °C, indicating pyrrhotite (Fig. 2c inset). For samples containing two phases, magnetite and pyrrhotite, soft and medium fractions prevail (Figs 2a and b insets). The hard component is carried by pyrrhotite and by magnetite or haematite, whereas the medium component is carried mainly by pyrrhotite. This suggests that pyrrhotite is harder (due to its fine grains or stress) when associated with magnetite. Clark (1984) found that small grains of pyrrhotite can be magnetically hard with a coercive force of about of 92 mT.

The IRM acquisition curves show the presence of both soft and hard magnetic phases for samples containing two magnetic minerals, pyrrhotite and magnetite (Fig. 3a, samples PB42 and ZB62). For rocks containing only pyrrhotite, only the soft phase is seen on IRM acquisition curves (Fig. 3a, sample PZ99). ARM acquisition curves (Fig. 3b) demonstrate two component behaviour for rocks with two magnetic carriers (samples PB42 and ZB62). One component saturated in a 50–70 mT peak alternating field whereas the other did not saturate in a field of 100 mT. Samples containing pyrrhotite show saturation in 50 and 100 mT fields (sample PZ99), whereas samples containing magnetite do not saturate in this field (sample DE31). Such behaviour of magnetite can be explained by the strong dependence of ARM intensity on grain size and field intensity (Dunlop & Argyle 1997; Jackson et al. 1988).

Hysteresis loop measurements, although obtained for a few specimens only, show remarkable differences in hysteresis parameters for rocks with a single magnetic mineral in comparison to rocks with two magnetic minerals (Table 1). For magnetite content, hysteresis parameters fall into the single domain area (after Dunlop & Özdemir 1997). For pyrrhotite, hysteresis parameters indicate a grain size of 25–15 µm (Dekkers 1988). Rocks with composite magnetic mineralogy show low coercive force and very high coercivity. We interpret this high coercivity as being due to the contribution of hard pyrrhotite with a grain size of less than 5 µm (Clark 1984; Dekkers 1988).

Palaeomagnetic experiments

Palaeomagnetic measurements were performed at three laboratories: the Palaeomagnetic Laboratory of the Institute of Geophysics in Warsaw, Geosciences Rennes at the University of Rennes 1, and the Institut de Physique du Globe at the University of Strasbourg. Natural remanent magnetization (NRM) was measured by means of 2G SQUID, LETI cryogenic, Schonstedt and Digico magnetometers. Demagnetization experiments were performed using an alternating field (AF) and thermal treatments. The results from all three laboratories are consistent. Chemical changes often did not allow for heating to temperatures higher than 450–500 °C. In some samples chemical alteration started at only 400 °C. Examples of thermal and AF demagnetization results are shown in Fig. 4. Characteristic remanence components were calculated using the PDA package (Lewandowski et al. 1997), which includes principal component analysis (Kirschvink 1980). Linear segments of demagnetization curves were calculated for maximum angular deviation (MAD) of less than 10°. More than four points were used to define a line. In the majority of specimens one component was determined for one specimen; however, for hand samples usually two components were observed. Within each site two groups of remanence were recognized, except two sites, M2 and DE, where three groups were found. The first group comprises directions (labelled R) removed between 0 and 20 mT or between 20 and 300 °C. They have in situ intermediate positive inclinations and northwesterly or northeasterly declinations (Figs 4e and f and 5). In site DE, a component labelled P was isolated that unblocked at temperatures of 325 or 450–500 °C and was removed by an alternating field of 20–30 mT (Figs 4c and d and 5). A direction labelled C of the same stability as the P direction was found in site PB. Stereographic projections of components isolated in particular sites are shown in Fig. 5.

For all sites a more stable component labelled D was removed at temperatures of about 325 °C or 450–500 °C and by a field of about 50–60 mT or higher (Figs 4a and b and 5). Unblocking temperatures of about 325 °C suggest that pyrrhotite is the carrier of remanence. Unblocking temperatures of about 450–500 °C indicate magnetite as the carrier of remanence. For sites PB and ZB, the D directions are distributed on the both sites of the equator and it is not obvious whether they are symmetrically distributed around the mean. Because of this, graphical displays were used for testing whether the Fisherian distribution is a suitable model for calculation of the site mean direction for the D component of these sites. Graphical displays were performed using a computer program package for analysis and presentation of palaeomagnetic data (Enkin 1994). Examples for sites PB and ZB are shown in Fig. 6. The data were plotted in such a way that the mean direction is the pole of stereographic projection (Figs 6a and b). Then, colatitude (Figs 6c and d) and longitude (Figs 6e and f) plots (Fisher et al. 1987) were made. Figs 6(a) and b show an approximately uniform distribution of directions around the mean. The colatitude plot is made using X_i = 1 − cos(90° − I′_i), where i = 1,…, N and I′_i is the inclination of the i direction rotated to the pole equal to the site mean direction. The colatitude plots (Figs 6c and d) should be approximately linear and pass through the origin, which is fulfilled. The slope of the plot gives an estimate of k. For sites PB and ZB estimates of k taken from the plots are 15.8 and 16.6, respectively, which is in good agreement with the k values of 13.3 and 16.2 calculated from the Fisherian distribution. The longitude plot is made using X_i = D′_i, where i = 1,…, N and D′_i is the declination of the i direction rotated to the pole equal to the site mean direction. The longitude plot should be linear, passing through the origin with slope 45°, which is also fulfilled (Figs 6e and f) for both sites.

For site PZ the distribution of the D directions is elliptical, not Fisherian; therefore, the Bingham distribution (Onstott 1980) was used and compared with the Fisherian statistics. The Bingham distribution gave the same value for the mean direction as the Fisherian distribution (D = 71°, I = 11°, α₃₂ = 13.04, α₃₁ = 5.25), so we stay with Fisher statistics for this site also.

Interpretation

The site mean directions are listed in Table 2 and shown in Fig. 7. The in situ P direction, I = − 16°, D = 208°, is close to the expected Permian direction calculated for the Sudetes from the Permian segment of the APWP, and corresponds to the palaeopole position (Fig. 8) consistent with the late Permian pole for Baltica (Torsvik et al. 1992). The P direction is found in two hand samples from one site only. Although we suspect that the P direction is a Permian remagnetization, it is not sufficiently well documented for interpretation. The in situ C direction (I = 25°, D = 216°) yields a palaeopole position slightly different from the Late Visean palaeopole for Baltica, latitude 25°S, longitude 339°E (Table 2, Fig. 8). After tectonic correction, the C palaeopole differs from the Devonian palaeopole and those younger then the Devonian palaeopole. The C direction is probably a Carboniferous overprint; however, the dip of olistoliths is meaningless for the magnetization acquired in the Carboniferous. The dip of the formation including olistoliths should be used. Unfortunately, it was not possible to measure such a dip. However, as in the case of the P direction, the C direction is not well documented either.

The in situ site means of the D directions are dispersed in declinations and inclinations (Fig. 7a). However, after tilt correction the declinations remain dispersed (Fig. 7b) but the inclinations became coherent. The D palaeopoles for sites ZB and PZ correspond in position to the Devonian segment of the APWP for Baltica (Fig. 8). The D palaeopole for site DE differs slightly from this segment whereas the D poles for sites M1, M2, M3 and PB lie far away from the path. The fold test for inclination only was applied to seven sites using the program of Enkin (1994) according to the method described by Enkin & Watson (1996). Such a test was applied by Parés et al. (1994) to examine oroclinal rotations in the Cantabrian/Asturian arc in Spain, and by Tarduno et al. (1990), who performed the ‘megaconglomerate test’ for nine blocks of pelagic limestones found within the Franciscan Central Belt melange of northern California. The test is positive at the 99 per cent confidence level with k₂/k₁ = 34.5, where k₁ is the value before untilting and k₂ is the maximum value of k (k₂ = 114, α₉₅ = 4°, I = − 20.8°) obtained for 100 per cent untilting (Fig. 9). This suggests that the studied upper Devonian rocks were magnetized before they entered the early Carboniferous turbiditic currents. After bedding correction, the D directions (Fig. 10) gave palaeolatitudes between 7°S and 15°S, which compare well with the palaeolatitudes expected for the GB from the APWP for Baltica (Torsvik et al. 1992), and fit the time interval of 375–390 Ma of the middle–late Devonian. The palaeolatitudes obtained also agree well with the value of 13°S obtained by Bachtadse et al. (1983) for Central Europe and with the value of 7.6°S obtained for the Late Devonian pole from Spitsbergen (Jeleńska & Lewandowski 1986). An Early Devonian palaeolatitude of 18.6°S obtained from the Ukrainian pole position also fits the data obtained here (Smethurst & Khramov 1992).

The site means of the R direction, in situ and after tilt correction, although well defined, are significantly different from any geomagnetic direction expected for the Sudetes from the APWP for Baltica (Figs 7c and d). After tectonic correction these directions are more dispersed than in situ directions (Fig. 7d). Looking for an interpretation we constructed for each site great circles from the R,D in situ directions and Permian (P_ex) or Carboniferous (C_ex) directions expected from the APWP for Baltica. For sites M1, M3, PZ and ZB, a great circle was determined from R, D and P_ex directions. For site M2, where two R components were observed, one great circle was drawn for the D, P_ex and NE R direction and a second great circle was drawn for the D, C_ex and NW R direction (Figs 11a and b). In Permian and Carboniferous times remagnetization was a strong and widespread phenomenon in the Sudetes. For site DE, the R component does not fit either the great circle including P_ex or the great circle with C_ex. The R component could be explained as an intermediate direction between normal and reversed D directions. The D direction for site DE has reversed polarity opposite to the normal polarity for sites M2, M3, PZ, PB and ZB. Reversed polarity was observed again in site M1. Unfortunately, we are not able to define the stratigraphic sequence of the sites but there is some field evidence that site M1 is younger than M2 and M3. On the palaeolatitude versus age path (Fig. 10), the sequence from younger to older is as follows: reversed polarity for M1, normal polarity for ZB, M2, M3, PB and PZ, and again reversed polarity for DE. Magnetostratigraphy data for times earlier than the Carboniferous are scarce and uncertain. For the Upper Devonian, Hurlay & Van der Voo (1990) reported mixed polarity, whereas Khramov (1982) found mostly reversed polarity for the Siberian platform, with the exception of short zone of normal polarity in the Frasnian. This is in agreement with our results.

Magnetic anisotropy

The anisotropy of magnetic susceptibility (AMS) was measured with a KLY 2 bridge and calculated with Jelinek's program Aniso 11 (Jelinek 1977). For each site, the data were subjected to eigenvector analysis using the program Spheristat 2.0 (Stesky 1995). Mean susceptibility ranged between 10⁻⁴ and 10⁻³ SI and is carried predominantly by paramagnetic minerals. The degree of anisotropy ranges between 1.03 and 1.12; for samples from the Podzamek limestones (site PZ) it even reaches values of 1.8. Such high values may be related to the presence of pyrrhotite. The shape parameter, T, usually has positive values indicating that foliation prevails. The site means of k_min after bedding correction show steep inclinations for all sites but sites M1 and PB (Fig. 12), indicating that primary sedimentary fabric could be preserved. The site mean k_max axes, both in situ and after correction for bedding, although well grouped within the site, show two almost perpendicular subhorizontal trends: NNW–SSE and ENE–WSW (Fig. 12). The NNW–SSE distribution of k_max axes observed in M1, M2 and M3 could be caused by significant rotation of these olistoliths compared to ZB, DE and PZ where the rotation is much smaller. Alignment of k_max for M1, M2 and M3 in an ENE–WNW direction brings D directions close to the value expected for the Devonian.

Discussion

The rock magnetic study showed that the main minerals responsible for the magnetic properties of the greywackes and mudstones are pyrrhotite and magnetite. In limestones pyrrhotite is the only magnetic carrier. Our palaeomagnetic study revealed four components of NRM (D, C, P and R) carried by pyrrhotite or sometimes by magnetite. The positive fold test for inclination only indicates the pre late Visean age of the D magnetization. It suggests that at least part of the pyrrhotite was produced at an early stage of diagenesis and acquired stable chemical remanent magnetization (CRM) at that time. According to Roberts & Turner (1993) and Berner (1970), pyrrhotite can be formed at the early stage of diagenesis and can be a carrier of magnetization acquired just after the formation of a rock. Menyeh & O'Reilly (1996) observed that fine particles of synthetic monoclinic pyrrhotite acquired an intense and stable TRM, so we can expect the same behaviour of CRM. The P and C components, although found in a few samples only and not well defined, could be regarded as Permian and Carboniferous overprints, respectively. As the R component lies on the great circle between the D component and P or C expected directions, we suggest that it resulted from Permian or Carboniferous partial remagnetization. The R component for site DE could represent an intermediate direction between reversed and normal Devonian directions.

We propose the following scenario of events for the ZBM complex.

• 1

  Devonian sedimentation took place and early diagenetic D magnetization was acquired.

• 2

  Early Carboniferous mass movements produced randomly oriented olistoliths and slide sheets within flysch. As a result, large rotations of the D component found in sites M1, M2, M3 (clockwise) and PB (counterclockwise) are observed.

• 3

  In Carboniferous times remagnetization of the ZBM complex took place. At site PB remagnetization resulted in the acquisition of the C direction. The acquisition of the Early Carboniferous direction between 350 and 340 Ma is consistent with the age of the Late Visean deformation of the ZBM complex (Wajsprych 1995). At site M2 partial remagnetization occurred.

• 4

  In the Permian, remagnetization took place—the in situ P component was acquired at the DE site whereas partial remagnetization took place at sites M1, M2, M3, PZ and ZB.

We should point out that we do not observe any differences in magnetic mineralogy between samples remagnetized in the Permian or Carboniferous and those that preserve the Devonian magnetization. Also, the unblocking temperatures that removed the DCP components are approximately the same. Only the unblocking values of the alternating field are usually higher for the D component than for the C and P components.

If we accept the Devonian origin of the D poles, our results indicate that the upper Ordovician–Devonian rocks of the Zdanów Bardo Młynów complex of the Góry Bardzkie must have been palaeogeographically close to or directly connected with Baltica during Middle–Late Devonian time, or at least they were positioned at the same latitudes. This suggests that the Rhenohercynian basin, which re opened north of the Middle German Crystalline High in Early–Middle Devonian times (Franke & Oncken 1995), was a narrow and short lasting event. It must have been even less significant further eastwards, to the north and east of the West Sudetes area, or its width there was beyond the resolution of palaeomagnetic methods. In the West Sudetes the Saxothuringian plate was already welded to the Tepla Barrandian terrane. Our data do not demonstrate any significant rotation of the West Sudetes after Devonian like that observed by Tait et al. (1996) in the Moravo Silesian basin. This throws new light on the supposed direct continuation of the Rhenohercynian basin to the Moravo Silesian basin of the East Sudetes.

Acknowledgments

This work was supported by the project CNRS MDRI (France)/Academy of Sciences (Poland) and by project 5/95 of the Institute of Geophysics (Polish Academy of Sciences). We are indebted to Dr Bernard Henry and Dr Robert Scholger for their useful comments.

References