The orthogneisses of the Orlica–Śniez2nik complex (West Sudetes, Poland): geochemical characteristics, the importance of pre-Variscan migmatization and constraints on the cooling history

Geochemical analyses in combination with Rb–Sr (whole-rock, phengite, biotite) and U–Pb zircon ages provide important constraints on the magmatic, metamorphic and structural evolution of the Śniez2nik and Gierałtów gneisses from the Orlica–Śniez2nik complex (West Sudetes). These two gneisses have been considered to represent distinct petrogenetic units; however, their major and trace element compositions as well as their Sr–Nd isotope characteristics show no systematic differences that are indicative for different protoliths. This striking similarity leads to the conclusion that the petrographic variability is caused by modifications superimposed during deformation and migmatization. Most εNd500 values are in the range between −3.3 and −5.7 and suggest derivation of the protoliths from pre-existing continental crust. Two-stage TDM model ages mostly fall in the range between 1.4 and 1.6 Ga and closely correspond to other orthogneiss occurrences in the Bohemian Massif. The Rb–Sr whole-rock system is disturbed on a regional scale, to variable degrees, resulting in dates (c. 450 Ma and c. 395 Ma) considerably younger than the time of magmatic crystallization (c. 500 Ma). Secondary ionization mass spectrometry (SIMS) U–Pb analyses provide two groups of 206Pb/238U ages (364–341 Ma and 527–472 Ma), which largely correlate with previously established ages for protolith formation and Variscan high-temperature metamorphism. No geochronological evidence for pre-Variscan (‘Caledonian’) events was found. For phengite and biotite, the Rb–Sr system yields ages of c. 340–320 Ma, which provide further constraints on the regional cooling history.

The Orlica-Ś nieżnik complex at the NE margin of the Bohemian Massif ( Fig. 1) represents one of the major lithostratigraphic units of the West Sudetes. This complex is mainly composed of high-grade amphibolite-facies orthogneisses and a varied metasediment series, which includes minor metavolcanic rocks (e.g. Don et al. 1990;Ż elaźniewicz et al. 2002). In addition, small bodies of eclogites and/or high-pressure (HP) granulites occur closely associated with orthogneisses in its NE part (e.g. Smulikowski 1967;Bakun-Czubarow 1991a, b, 1992Bröcker & Klemd 1996;Kryza et al. 1996). The tectonometamorphic evolution of the Orlica-Ś nieżnik complex is difficult to unravel, because of complex field relationships and problems in dating distinct stages of an apparently polyphase P-T-t-deformation path. There seems to be a general consensus that most orthogneisses have Cambrian-Ordovician protolith ages (c. 520-490 Ma; e.g. Oliver et al. 1993;Borkowska & Dörr 1998;Kröner et al. 2001;Š típská et al. 2004) and that all rock types underwent at least one episode of medium-to high-temperature metamorphism during Variscan times (c. 340 Ma; e.g. Borkowska et al. 1990;Turniak et al. 2000;Lange et al. 2002). Although geochronological results are practically indistinguishable (c. 340 Ma, e.g. Brueckner et al. 1991;Klemd & Bröcker 1999), textural relationships suggest that an eclogite-facies stage preceded amphibolite-facies metamorphism. The importance of pre-Variscan ('Caledonian') orogenic events for the metamorphic and deformational history of the Sudetes is not unambiguously documented and is a matter of considerable debate (e.g. Don 1990;Oliver et al. 1993;Johnston et al. 1994;Cymerman et al. 1997;Bröcker et al. 1998;Kröner & Hegner 1998;Aleksandrowski et al. 2000;Kröner et al. 2001;Ż elaźniewicz et al. 2002). For the western Czech part of the Orlica-Ś nieżnik complex (¼ Orlické hory), pre-Variscan deformational and metamorphic processes were suggested by Přikryl et al. (1996), and this model seems to be supported by 207 Pb/ 206 Pb zircon evaporation dating, indicating that the foliation in the orthogneisses is older than a c. 492 Ma granodiorite vein (Kröner et al. 2001). In the eastern Polish part of the Orlica-Ś nieżnik complex, field observations indicate two generations of migmatites (Franke & Ż elaźniewicz 2000;Ż elaźniewicz et al. 2002). Locally, augengneisses contain isolated enclaves of migmatized rocks, which were included in the granitic magma during emplacement (Grześkowiak & Ż elaźniewicz 2002). This relationship has been interpreted to indicate that granite intrusions were coeval with or slightly younger than early migmatization (Franke & Ż elaźniewicz 2000;Grześkowiak & Ż elaźniewicz 2002). At other places, migmatitic rocks apparently developed at the expense of the augen-gneisses (e.g. Don 2001). It is unclear whether the latter type records a single process or documents multiple episodes of migmatization, which possibly are significantly separated in time. In addition, the time of the main deformation (pre-Variscan v. Variscan age) is under discussion (e.g. Don et al. 1990;Ż elaźniewicz et al. 2002).
Further complications for understanding the evolution of the Orlica-Ś nieżnik complex arise from ambiguous genetic relationships between two orthogneiss varieties, the Ś nieżnik and Gierałtów gneisses (e.g. Don et al. 1990;Ż elaźniewicz et al. 2002). Unsolved problems include the importance of textural, bulk and mineral chemical differences between these two gneiss types (e.g. Borkowska et al. 1990;Don et al. 1990;Borkowska & Dörr 1998;Turniak et al. 2000;Don 2001;Kröner et al. 2001;Grześkowiak & Ż elaźniewicz 2002;Lange et al. 2002). It has proven difficult to reconcile the contrasting views about the origin of the two gneiss varieties that are based on field observations and previously reported geochemical data.
This study is a contribution to the debate about the magmatic and tectonometamorphic history of the Orlica-Ś nieżnik complex and focuses on two problematic aspects: the genetic relationship between different types of orthogneisses and the significance of pre-Variscan metamorphism. Using conventional geochemical analyses (including REE), Sr-Nd isotope data and mineral chemistry, compositional differences between the Ś nieżnik and Gierałtów gneisses are re-examined. Rb-Sr whole-rock and U-Pb zircon dating is used to constrain the age of high-grade metamorphism and to find indications for the suggested multiple migmatization events. Furthermore, the timing of the postorogenic cooling history is addressed using Rb-Sr phengite and biotite chronology.

Geological setting
The Orlica-Ś nieżnik complex outcrops in the West Sudetes are located between two NW-trending fault zones, the Sudetic Marginal Fault in the NE and the Bušin Fault in the SW (Fig. 1). Both fault zones extend parallel to the Odra and Elbe Fracture Zones, the major lineaments dividing the West Sudetes from other crustal segments within the Variscides. The eastern border is defined by the Moldanubian Thrust Zone along which the Orlica-Ś nieżnik complex is in tectonic contact with the Moravo-Silesian footwall. In the west, the Olešnice-Uhřinov Fault separates the Orlica-Ś nieżnik complex from the lower-grade Nové Město phyllites. This feature combined with opposite dips has been traditionally, although not fully consistently, interpreted to represent a domal structure (e.g. Don et al. 1990). A metagranitic to migmatitic core, including tectonic bodies of mafic and acidic HP to UHP rocks, is mantled by metasediments and metabasic rocks, metamorphosed under middle-to lower amphibolite-facies P-T conditions, and an outer envelope of similar lithologies, with mineral assemblages testifying to greenschist-facies metamorphism (e.g. Ż elaźniewicz et al. 2002). Orthogneisses are the dominant rock type in the core and traditionally are subdivided into two varieties: the variably deformed, coarse-grained Ś nieżnik augen-gneisses and the finegrained migmatitic Gierałtów gneisses (e.g. Don et al. 1990;Ż elaźniewicz et al. 2002). The timing of emplacement of the granitic precursors is broadly constrained to between c. 520 and  (e.g. van Breemen et al. 1982;Oliver et al. 1993;Turniak et al. 2000;Kröner et al. 2001;Š típská et al. 2004).
Variable degrees of migmatization in the gneisses and staurolite-bearing mineral assemblages in the Stronie metasediments indicate that upper amphibolite-facies P-T conditions were attained in most parts of the Orlica-Ś nieżnik complex. Some workers concluded that the HP metamorphism recorded in eclogites and granulites also affected the country rock gneisses (e.g. Brueckner et al. 1991;Bröcker & Klemd 1996). Other interpretations suggest that the HP rocks were enclosed in a migmatitic gneiss diapir Don 2001), ductilely extruded from the lower crust (Š típská et al. 2004), or tectonically inserted into their present surroundings along localized subvertical shear zones (e.g. Stawikowski 2002;Ż elaźniewicz & Bakun-Czubarow 2002). For eclogites and granulites, estimated peak pressures lie in the coesite stability field (P .27 kbar) at temperatures in the range of 700-800 8C and 800-1000 8C, respectively (Bakun-Czubarow 1991a, b;Bröcker & Klemd 1996;Kryza et al. 1996;Klemd & Bröcker 1999). On the other hand, the Młynowiec-Stronie metasediments and associated basic and acid metavolcanic rocks were metamorphosed at peak temperatures of c. 620-650 8C and maximum pressure <8 kbar (e.g. Jastrzȩbski 2003;Murtezi 2003). The deformation history of the Orlica-Ś nieżnik complex is complex and comprises at least three stages. In some studies the main deformation (D 2 and D 3 ) is interpreted as a Late Caledonian process, whereas in others the importance of Early to Late Carboniferous Variscan events is emphasized (for overviews see Don et al. 1990;Ż elaźniewicz et al. 2002).
Previous geochronology has documented the importance of the Variscan orogeny for the tectonometamorphic evolution of the study area. Rb-Sr and 40 Ar/ 39 Ar dating mostly provided Carboniferous cooling ages for orthogneisses (c. 340-330 Ma;Borkowska et al. 1990;Steltenpohl et al. 1993;Lange et al. 2002;Marheine et al. 2002). A similar age was obtained by sensitive high-resolution ion microprobe (SHRIMP) U-Pb dating of zircon overgrowths (342 AE 6 Ma; Turniak et al. 2000), interpreted to closely approximate the time of HT metamorphism. Sm-Nd and U-Pb ages for eclogites and HP-granulites mostly cluster at c. 340 Ma (Brueckner et al. 1991;Klemd & Bröcker 1999;Š típská et al. 2004), but recently Szczepanski et al. (2004) also reported for granulites a Lu-Hf garnet age of c. 380 Ma.

Samples and methods
To ensure a representative regional characterization and to complement existing datasets, we have systematically collected 15 Ś nieżnik gneiss and 24 Gierałtów gneiss samples across the Polish part of the Orlica-Ś nieżnik complex for conventional geochemical analyses (including REE) and Sr-Nd isotope studies (including Rb-Sr and Sm-Nd geochronology). Our results represent the most comprehensive geochemical and isotope geochemical database for orthogneisses of the Orlica-Ś nieżnik complex. For U-Pb zircon dating, we have selected an intrusive anatectic mobilizate (sample 115) within migmatitic Gierałtów gneisses. Petrographic descriptions of the orthogneisses have been given by Don et al. (1990) and Lange et al. (2002).
Photographs of representative hand-specimens, sample location maps, details of the analytical methods as well as mineral compositional data and related figures are available online at http://www.geolsoc.org.uk/ SUP18224. A hard copy can be obtained from the Society Library.

Bulk-rock compositions and Sr-Nd isotope characteristics
The Ś nieżnik and Gierałtów gneisses have granitic compositions. Both varieties are compositionally homogeneous and show rather limited variations in the concentrations of their major, minor and trace elements including REE (Fig. 2). The compositional ranges overlap and significant differences between the two groups cannot be recognized. Initial 87 Sr/ 86 Sr ratios, calculated for a presumed protolith age of 500 Ma, range from 0.6944 to 0.7074 for Ś nieżnik gneisses and from 0.6651 to 0.7088 for Gierałtów gneisses. The åNd 500 values of Gierałtów gneisses range from À0.3 to À7.1, but most of them lie between À3.3 and À5.7; åNd 500 for Ś nieżnik gneisses varies from À3.5 to À5.2. Model age calculations are based on the two-stage model of Liew & Hofmann (1988), taking into account Sm/Nd modifications caused by high-T metamorphism and melt fractionation in igneous rocks. The need for such a correction is indicated by a large scatter of data points in a Sm-Nd isochron diagram (not shown here), interpreted to document that Sm and Nd did not fractionate coherently between melt and restite. Two-stage T DM ages (¼ depleted mantle model ages) for both groups of orthogneisses cluster between 1.4 and 1.6 Ga (Table 1), but Gierałtów gneisses show a larger range (1.2-1.7 Ga) than Ś nieżnik gneisses (1.5-1.6 Ga). The new åNd 500 values and T DM ages (n ¼ 39) agree very well with the smaller dataset for the westernmost Orlica-Ś nieżnik complex of Hegner & Kröner (2000) and Kröner et al. (2001) (n ¼ 9). Those workers reported åNd(t) values ranging from À3.5 to À6.5 and T DM ages between 1.4 and 1.7 Ga for granitic gneisses.
To further constrain the cooling history on a regional scale, we have selected 12 samples for Rb-Sr phengite and biotite dating ( Table 2). The samples were collected around the locations Nowa Wieś, Miȩdzygórze, Idzików, Nowa Morawa, Stronie Ś laskie, Strachocin and Ladek-Zdrój. According to previous classifications, seven samples represent Gierałtów gneisses and five samples belong to Ś nieżnik -type gneisses. 87 Rb/ 86 Sr ratios of phengite and biotite generally are high (c. 44-1583; Table 2), suggesting that the isochron ages are well constrained. All age calculations are based on mica-whole-rock pairs. Phengite from Gierałtów gneisses yielded ages between 339.9 AE 3.8 and 330.4 AE 3.5 Ma (weighted average 334.9 AE 4.3 Ma, n ¼ 5). For biotite of this rock group, ages between 336.9 AE 3.4 Ma and 318.7 AE 3.2 Ma were obtained. Four out of six samples yielded a weighted average of 331.3 AE 6.5 Ma, whereas significantly younger ages of 318.9 AE 3.2 Ma and 318.7 AE 3.2 Ma were obtained for the two other samples. Ś nieżnik gneisses yielded phengite ages between 341.6 AE 3.7 Ma and 333.5 AE 3.5 Ma (weighted average 337.3 AE 3.9 Ma, n ¼ 5). The corresponding biotite ages range from 333.8 AE 3.4 Ma to 330.4 AE 3.4 Ma (weighted average 331.9 AE 1.7 Ma, n ¼ 4). It is important to record that no differences in phengite and biotite ages were recognized between the two orthogneiss varieties. In some samples phengite and biotite ages are slightly different; in other cases biotite ages are younger. This seems to indicate a trend for phengite being older than biotite (Table 2), but this is not well resolvable, because the different ages overlap within analytical uncertainties. Across the study area, no regional difference in mica ages can be recognized.

U-Pb zircon dating
Sample 115, collected close to Strachocin, represents an anatectic mobilizate in migmatitic Gierałtów gneisses. Zircons from this sample are generally lightly coloured and euhedral to subhedral with short-prismatic to normal-prismatic morphology (typical elongation 2:1 to 3:1). Internally, they are very complex, with several generations of zircon growth. Most grains comprise a core surrounded by a euhedral overgrowth of variable width. The cores usually make up about 80-90% of the grains by volume, although in rare cases the overgrowths predominate. The cores show well-developed oscillatory zoning indicating an igneous (felsic) source. The overgrowths form mainly as euhedral pyramidal tips, which appear dark under CL imaging (equating  to relatively high U contents) with little, or only faint, broad zoning. The core-rim boundary is usually marked by a zone of lighter coloured zircon that probably represents a reaction zone or recrystallization front. In other instances the overgrowths clearly erode and embay the older protolith zircons. An additional complication is that the internal structure of these zircons contain small rounded cores within the magmatic protolith zircons.
Zircons selected for single-grain isotope dilution thermal ionization mass spectrometry (ID-TIMS) analyses were clear and without visible inclusions or microfractures. All grains were airabraded before dissolution, to remove the outermost surface. Judging from the CL images, it was expected that in most cases (1) the zircon overgrowths could be eliminated completely by air-abrasion, (2) only a small number of grains would be affected by inheritance, and (3) such grains could easily be identified, because of obviously discordant dates. U-Pb results are reported in Table 3 and Secondary ionization mass spectrometry (SIMS) U-Pb dating was carried out by use of a sensitive high-resolution ion microprobe (SHRIMP). The data for 24 SHRIMP analyses that were performed on 20 zircons are presented in a conventional U-Pb Wetherill concordia plot ( Fig. 5; Table 4). Site selection for analysis was based on CL images. Even with the aid of postanalysis high-magnification SEM and CL imaging it could not completely be confirmed that all selected spots were exclusively located in one specific growth zone. Figure 5 shows the two main populations of zircon representing the zoned inner parts and the overgrowths. These zircons record a complex history.

Discussion
The Śnieżnik and Gierałtów gneisses: genetic relationships and source characteristics The petrographically based subdivision of the orthogneisses into at least two groups gave rise to a long-lasting debate on whether these rocks are derived from different protoliths or from identical  Fig. 4). However, as discussed below, ages deduced from Rb-Sr isochron diagrams are younger than the true age of emplacement. Consequently, neither the dates nor the 87 Sr/ 86 Sr(t) values can reliably be assigned any geological relevance. Previously reported mineral chemical differences between the  Errors are 1ó; Pb c and Pb* indicate the common and radiogenic portions, respectively. Isotopic ratios are common Pb corrected using measured 204 Pb. two gneiss varieties (e.g. Borkowska & Dörr 1998;Borkowska & Orłowski 2000) are confirmed by new data for biotite and plagioclase (Lange 2004). Because no bulk-compositional variations were detected between the two gneiss varieties, these differences are probably related to compositional modifications during metamorphic overprinting (migmatization). The results of this study are compatible with the interpretation that all orthogneisses of the Orlica-Ś nieżnik complex belong to a single igneous suite, derived from identical source rocks. It is suggested that the petrographic and mineral compositional variability was mainly caused by modifications superimposed during deformation and migmatization. The Ś nieżnik and Gierałtów gneisses represent different textural variants of the same protolith, as was also suggested by Turniak et al. (2000) and Kröner et al. (2001). This conclusion is, however, not fully supported by the observation of Gierałtów-type, migmatitic xenoliths within Ś nieżnik augen-gneiss (Grześkowiak & Ż elaźniewicz 2002), unless the captured fragments represent parent or country rock that underwent synintrusive or earlier migmatization.
The range in zircon ages (c. 520-490 Ma) is in accord with a model suggesting that compositionally uniform granitic magmas intruded over an extended period of time, and that the Orlica-Ś nieżnik complex orthogneisses represent a large batholith composed of different magma pulses. However, it cannot completely be ruled out that this range is an artefact of variable degrees of undetected inheritance and/or disturbance of the U-Pb system. In any case, åNd(t) values (mostly À3.3 to À5.7), T DM ages (mostly between 1.4 and 1.6 Ga) and the presence of inherited zircons suggest that melts were extracted from a source dominated by old continental crust (Turniak et al. 2000;Kröner et al. 2001;this study). However, one Gierałtów gneiss (sample 166) is characterized by a considerably lower åNd 500 value of À0.3 (T DM ¼ 1:2Ga), indicating mixing with juvenile source material; another orthogneiss of this rock group (sample 103) also shows a slight deviation from the main trend (åNd 500 À2.5), pointing to a smaller degree of mixing with mantle-derived material. T DM ages of granitic gneisses from the Orlica-Ś nieżnik complex mostly fall within the range observed elsewhere at the NE margin of the Bohemian Massif (Erzgebirge, Lusatian Block, Góry Sowie, Lugian Domain; for a compilation of the relevant data see Hegner & Kröner 2000). Although the database still is very small, U-Pb SHRIMP and 207 Pb/ 206 Pb evaporation dating of zircon xenocrysts suggest derivation from a heterogeneous basement, composed of various crustal components with minimum 207 Pb/ 206 Pb ages ranging from c. 530 Ma to c. 2.6 Ga (Turniak et al. 2000;Kröner et al. 2001;this study). Because of this heterogeneity and complexities potentially arising from variable degrees of mixing, including juvenile material, caution is warranted in using T DM ages for palaeogeographical linkages (Crowley et al. 2002).
The main purpose of the geochronological studies presented here was to unravel the importance of pre-Variscan HT metamorphism for the study area. Thus, the variably migmatized Gierałtów gneisses were the main target of this study. Rb-Sr whole-rock dating has the potential to look through isotopic resetting processes, which are effective on the mineral scale. In favourable circumstances, this method can provide insights into the earlier stages of the magmatic or metamorphic history. By use of Rb-Sr whole-rock dating we obtained an isochron age of 449 AE 5 Ma for the Gierałtów gneisses, which is significantly younger than the inferred time of protolith formation (c. 520-490 Ma;e.g. van Breemen et al. 1982;Liew & Hofmann 1988;Oliver et al. 1993;Turniak et al. 2000;Kröner et al. 2001;Š típská et al. 2004). In combination with textural characteristics of this rock group, clearly testifying to anatectic processes, the Rb-Sr data might be interpreted as an indication for pre-Variscan migmatization. This hypothesis is apparently supported by U-Pb ages of zircons derived from a leucocratic melt injection (sample 115) within Gierałtów gneisses. Single-grain ID-TIMS dating yielded an upper intercept age of 446 AE 8 Ma ( Fig. 4b; Table 3), which overlaps with the Rb-Sr whole-rock age for Gierałtów gneisses. However, CL imaging revealed a heterogeneous zircon population with variable degrees of internal complexities. Grains with different CL patterns could not be distinguished by morphological, colour or size criteria under the stereomicroscope. Therefore, despite air-abrasion, the grains selected for ID-TIMS analysis may have consisted of zones with different internal structures and ages, resulting in isotopic data without geological significance. SHRIMP U-Pb dating provided two age populations for central grain parts and overgrowths, clustering at c. 490 Ma and c. 350 Ma, respectively. The chemistry (very low Th/U) and the form of the overgrowths is consistent with growth under metamorphic conditions. The SHRIMP ages correlate perfectly with two well-documented geological events in the Orlica-Ś nieżnik complex history (i.e. protolith formation and Variscan HT metamorphism), and suggest that the corresponding Rb-Sr and U-Pb ID-TIMS dates of c. 450 Ma are fortuitous. SHRIMP dating clearly shows that all data points obtained by ID-TIMS single-grain dating represent geologically meaningless mixtures. The c. 450 Ma dates have no geological relevance and are related to disturbance of the Rb-Sr whole-rock system during metamorphic overprinting and mixing of zircon domains with different ages.
The Rb-Sr whole-rock system of the orthogneisses is disturbed to variable degrees on a regional scale. The considerably younger dates of c. 395 Ma reported for the Miȩdzygórze area (Borkowska et al. 1990;Bröcker et al. 1997; this study) indicate a more severe overprint than elsewhere in the Orlica-Ś nieżnik complex. On the other hand, the Rb-Sr whole-rock age of 487 AE 11 Ma age described by van Breemen et al. (1982) for the Ś nieżnik type gneisses from the Czech part of the Orlica-Ś nieżnik complex closely approximates the inferred time of magma emplacement. What is the reason for the disturbance of the Rb-Sr system? It cannot completely be ruled out that at least the c. 450 Ma date reflects the influence of pre-Variscan orogenic processes. This does not necessarily imply a migmatization event, but may indicate a distinct deformation stage. However, geochronological evidence for such processes has not been found yet. Judging only from the available chronological constraints, the most plausible interpretation is to suggest that incomplete rejuvenation of the Rb-Sr whole-rock system is related to Variscan metamorphism, dated at c. 340 Ma (e.g. Turniak et al. 2000;Lange et al. 2002).

Mica geochronology: constraints for the cooling history
The new phengite and biotite ages support previous interpretations based on a limited dataset from a single outcrop at Miȩdzygórze (Lange et al. 2002) and further document the importance of Variscan metamorphism in the western Sudetes. Phengite-and biotite-whole-rock pairs of Gierałtów gneisses yielded Rb-Sr ages of c. 340-330 Ma and c. 337-319 Ma, respectively. For Ś nieżnik gneisses, phengite ages range from c. 342-334 Ma and biotite ages from c. 334-330 Ma. Uncer-tainty on these ages generally is c. 1% (Table 2). Systematic regional variations in mica ages were not recognized. The Rb-Sr results are in perfect agreement with 40 Ar/ 39 Ar phengite and biotite cooling ages of 341 AE 1 to 337 AE 0.6 Ma and 342 AE 1 to 334 AE 0.6 Ma, respectively, reported for micaceous gneisses and schists from the eastern part of the study area, as well as 40 Ar/ 39 Ar ages between 338 AE 0.9 and 335.0 AE 0.5 Ma for the western Orlica-Ś nieżnik complex (Glascock et al. 2003).
Within individual samples, phengite and biotite ages often overlap within error, but phengite ages show a trend towards slightly older values. Because of their different closure temperatures for the Rb-Sr system, these differences are interpreted to indicate cooling after a thermal event. As underlined by the weighted averages, phengite and biotite of both orthogneiss varieties define relatively homogeneous age populations (phengite: 334.9 AE 4.3 Ma, n ¼ 5 and 337.3 AE 3.9 Ma, n ¼ 5; biotite: 331.3 AE 6.5 Ma, n ¼ 4 and 331.9 AE 1.7 Ma, n ¼ 4). A distinctly younger group of biotite ages (c. 319 Ma) was found in some Gierałtów gneisses and may be related to slower cooling in some rock domains, retrograde disturbance as a result of late fluidrock interaction, or discrete deformation events. The last interpretation is consistent with conclusions based on 40 Ar/ 39 Ar phengite and biotite dating, suggesting small-to large-scale shear movements in the Orlica-Ś nieżnik complex at around 325-320 Ma (Marheine et al. 2002). Even younger Rb-Sr biotite dates occur locally (c. 315-293 Ma; Lange et al. 2002;Bröcker et al., unpubl. data). Especially interesting are ages around 300 Ma, which were previously unknown from the Orlica-Ś nieżnik complex. Similar 40 Ar/ 39 Ar biotite and muscovite ages (c. 310-300 Ma) are widespread in the Keprník and Desná massif (Maluski et al. 1995), located to the SE of the study area, and were interpreted to date a major tectonometamorphic phase associated with extensional processes, which largely erased the isotopic record of earlier Variscan events (Maluski et al. 1995). This metamorphic episode apparently also affected the Orlica-Ś nieżnik complex in strongly localized shear zones.

Conclusions
The tectonometamorphic history of the Orlica-Ś nieżnik complex is complex and comprises at least one migmatization event at c. 340 Ma (e.g. Turniak et al. 2000;Lange et al. 2002;this study), but many aspects of the P-T-t-deformation path remain enigmatic. Our attempts to look through the Variscan overprint to date earlier stages were not successful. At present, field observations and the geochronological record cannot unambiguously be reconciled. In the studied rocks, the geochronological record of pre-Variscan metamorphism and deformation was completely erased during pervasive HT overprinting at c. 340 Ma. An alternative explanation, not favoured here as being in conflict with the field data, is to suggest that the orogenic history comprises only Variscan events.
Geochronological indications for pre-Variscan events were described from the Orlické hory in the westernmost Czech part of the Orlica-Ś nieżnik complex. From this area, Kröner et al. (2001) reported a 207 Pb/ 206 Pb zircon evaporation age of 491.7 AE 1.0 Ma for an undeformed amphibole-biotite microgranite dyke, which cuts foliated felsic orthogneisses, suggesting a minimum age of c. 492 Ma for the foliation. Those workers suggested that the Orlica-Ś nieżnik complex was affected shortly after magma emplacement by regional metamorphism accompanied by ductile shearing. At present, it cannot completely be ruled out that these zircons represent xenocrysts, as observed in our study for the anatectic mobilizate from Strachocin (see Tikhomirova 2002). Because of its importance for regional geological considerations, this date and its geological relevance needs to be tested by microbeam U-Pb techniques (SIMS, laser ICP-MS).
This study exemplifies the problems encountered in attempts to unravel the magmatic and P-T-t-deformation history of pervasively overprinted rocks that have undergone both polyphase deformation and possibly also multiple episodes of metamorphism. Further in-depth understanding of the tectonometamorphic evolution of the Orlica-Ś nieżnik complex can be obtained only if precise ages for chronologically yet unconstrained events become available. As a result of severe Variscan overprinting, the prospects for dating directly earlier tectonic features are rather small. A promising approach, both for indirectly dating deformation and for directly dating distinct metamorphic stages, is to focus further isotopic research on anatectic mobilizates, melt patches and leucosomes. The geochronological database for such rocks still is very small, comprising only three samples (Kröner et al. 2001;Š típská et al. 2004;this study). In our view, additional dating of similar rocks is the key to obtain a more comprehensive insight into the tectonometamorphic evolution of the Orlica-Ś nieżnik complex.
Thanks are due to H. Baier for laboratory assistance and support on the mass spectrometer. The constructive reviews by M. Tichomirowa and an anonymous reviewer are gratefully acknowledged. This study was supported by the Deutsche Forschungsgemeinschaft (grant BR 1068/7-1, 7-2).