Analytical techniques

Mineral compositions were determined using a CAMECA SX100 electron microprobe at the Institut für Mineralogie, Petrologie und Geochemie at the University of Freiburg, Germany, and a CAMECA CAMEBAX MICROBEAM with internal PAP correction (Pouchou & Pichoir, 1984) at the Institut für Geowissenschaften at the University of Mainz, Germany. Natural and synthetic standards were used for major and minor elements. Measuring times per element were 20 s with an emission current of 10 nA and an acceleration voltage of 15 kV. Na always was measured first to avoid errors resulting from the volatilization of Na.

X-ray fluorescence (XRF) analyses of whole-rock samples and mineral separates (plagioclase, clinopyroxene, orthopyroxene) were carried out at the Institut für Mineralogie, Petrologie und Geochemie at the University of Freiburg, Germany, on a Philips PW 1450/20 machine with natural standards; accuracy and detection limits are on the order of 0·1 wt % for major elements and 1-10 ppm (depending on the specific element) for minor elements. Preparation was performed at the same institute. For whole-rock measurements, ~1-3 kg of material, depending on the grain size, was crushed and milled in agate mills, and pressed powder and Li-borate fusion discs were prepared to measure trace and major elements, respectively. Mineral separates were prepared from 63-125 µm fractions of the whole-rock samples; separations were made using heavy liquids, magnetic separators and a wave table. The resulting separates were again milled, and pressed powder as well as fusion discs were prepared.

Two methods were used to obtain bulk compositions of exsolved magmatic pyroxenes: if the exsolution lamellae were small enough ( Fig. 4), the pyroxenes were analysed by electron microprobe using a defocused beam (between 10 and 50 µm) and thus reintegrated. Where the exsolution lamellae were on the order of 10-100 µm thick ( Fig. 5) the lamellae and the host were measured separately and the original pyroxenes were reintegrated using image processing techniques. Neither method successfully reintegrated magmatic compositions for exsolved feldspars and oxides, as the results showed wide scatter and gave low mineral equilibration temperatures indicating later re-equilibration.

Figure 4. Photomicrographs of pyroxene exsolution textures in mangeritic and charnockitic rocks. (a) Clinopyroxene from sample GM 254 with thin orthopyroxene exsolution lamellae; this type of exsolution was reintegrated with a defocused beam on the microprobe; (b) coarsely exsolved crystal of inverted pigeonite surrounded by mesoperthite in sample GM 254; (c) coarsely exsolved crystal of inverted pigeonite in mesoperthite, surrounded by a small rim of later garnet; the crystals in (b) and (c) could not be reintegrated reliably. Scale bar represents ~1 mm.

Figure 5. Photomicrographs of pyroxene exsolution textures in anorthositic rocks. (a) Narrow exsolution lamellae of orthopyroxene in a crystal of clinopyroxene in plagioclase matrix; this crystal comes from sample GM 263 and was reintegrated with image analysis (see text); (b) very finely exsolved crystal of orthopyroxene in sample GM 265; this crystal was reintegrated with a defocused beam on the microprobe; plagioclase in the lower right; (c) coarsely exsolved clinopyroxene crystal in parts exhibiting `blebby' exsolutions with surrounding plagioclase in sample GM 334; this crystal could not be reintegrated reliably. Scale bar represents ~1 mm.

Mineral abbreviations used in the text or in the figures are: cpx, clinopyroxene; opx, orthopyroxene; pig, pi- geonite; CaTs, Ca-tschermakite; ol, olivine; mt, magnetite; usp, ulvøspinel; mgsf, magnesioferrite; ilm, ilmenite; hem, haematite; plag, plagioclase; msp, mesoperthite; kfsp, potassium feldspar; zir, zircon; ap, apatite.

Olivine

Olivine ranges in composition from Fo₈₃ in the Moskenesøy cumulates to nearly pure fayalite in thecharnockites ( Table 2 and Fig. 6). In the basic rocks, olivine ranges from Fo₈₃ in the Moskenesøy cumulate to Fo₇₇ in the Ramberg gabbro on Flakstadøy (GM 349) and to Fo_72-66 in the anorthositic rocks of the FBC. CaO is below detection limit in all samples and MnO is on the order of 0·3 wt %. The Ni content decreases from 130-190 ppm in the cumulate olivines from Moskenesøy to 20-50 ppm in olivines from the anorthositic rocks. The jotunite GM 384 has Fo₁₉ olivine ( Table 2). The mangeritic rocks contain olivines between Fo₉ and Fo₅. Their MnO content ranges from 1·5 wt % in the jotunite to 2·6 wt % in the mangerites and charnockites.

Table 2. Selected electron microprobe analyses of olivine from the Lofoten Islands.

Figure 6. Reintegrated, measured and calculated magmatic pyroxene and olivine compositions from anorthositic, mangeritic and charnockitic rocks and from the cumulate on Moskenesøy plotted within the pyroxene quadrilateral after the method of Lindsley, (1983). (See text for explanation of reintegration and calculation methods and for discussion.)

Pyroxenes

As with olivine, the most magnesian augite is found in the Moskenesøy cumulates [X_Fe = Fe/(Fe + Mg) = 0·42]. Clinopyroxenes in the gabbroic and anorthositic rocks are more Fe rich (up to X_Fe = 0·5). Compositions of reintegrated subcalcic augites are given in Table 3 and plotted in Fig. 6, using the projection scheme of Lindsley, (1983). Generally, the augite has low contents of Na₂O and Cr₂O₃, but relatively high contents of Al₂O₃ (between 3·3 and 4·7 wt %). The Al₂O₃ content varies significantly between samples and to a small extent within individual samples ( Table 4). In individual rocks, megacrysts show the same range in Al₂O₃ as the small interstitial grains. The trace-element contents of a clinopyroxene megacryst mineral separate from the FBC (GM 520) are given in Table 5. Because of their Al₂O₃ contents and their X_Fe, the Lofoten megacrysts are similar to type II megacrysts reported by Emslie, (1975), which are interpreted to have crystallized within the anorthosite, as opposed to the very aluminous type I pyroxenes, which are interpreted to have crystallized at greater depths and to be xenocrysts (Emslie, 1975). In the jotunite, clinopyroxene has lower Al₂O₃ contents (~1·5 wt %) and in the mangerites, the Al₂O₃ content ranges from 0·5 to 0·9 wt %. There is a marked increase in X_Fe between the clinopyroxenes in the jotunite and the anorthositic rocks and a continuous increase from jotunite (X_Fe = 0·68) to mangerite to charnockite (X_Fe up to 0·96, see Table 3).

Table 3. Selected analyses of reintegrated clino- and orthopyroxenes (for the reintegration techniques and the sample numbers, see text and Table 1); also listed are the compositions in terms of end-members according to the En-Wo-Fs projection of Lindsley, (1983).

Table 4. Selected microprobe analyses of exsolved clinopyroxenes from anorthositic rocks of the FBC, Lofoten.

Table 5. XRF analyses of mineral separates from an ortho- and clinopyroxene megacryst from anorthositic rocks on the Lofoten Islands.

Orthopyroxene

Orthopyroxene from the anorthosites has X_Fe values ranging from ~0·26 (FBC) to 0·31 (Eidsfjord). The Al₂O₃ contents are around 2·7-3·2 wt % for both interstitial and megacrystic grains. The megacrysts are similar to Emslie's, (1975) type II megacrysts. The CaO content of reintegrated crystals is between 1·9 and 2·1 wt %. Table 5 gives major and trace elements of an orthopyroxene megacryst (GM 450) from the Eidsfjord anorthosite and Table 3 contains two reintegrated orthopyroxene analyses of interstitial grains from the FBC and the Eidsfjord anorthosite. The mangeritic rocks contain orthopyroxene with X_Fe values between 0·49 and 0·69. Figure 6 shows plots of reintegrated and calculated equilibrium orthopyroxenes.

Pigeonite

Inverted pigeonite with X_Fe values between 0·79 and 0·86 was found locally in a few mangerites and charnockites. There are no analyses given as it was not possible to reintegrate the exsolved inverted pigeonites. The pigeonite plotted in Fig. 6 was calculated from the constraints of coexisting augite using the QUILF program of Andersen et al., (1993).

Fe-Ti oxides

Most rocks contain both primary ilmenite and magnetite, although some charnockites lack magnetite. In the FBC, green hercynitic spinel is commonly observed as an exsolution phase from the primary spinel. In rocks containing both primary magnetite and ilmenite, extensive `oxy-exsolution' and Fe-Ti exchange precludes determination of the original compositions of either phase; they are therefore not reported. In some mangeritic and charnockitic rocks, only ilmenite was observed. Here, oxy-exsolution and exchange of Ti did not occur and their composition is regarded as primary. The measured compositions from selected samples are reported in Table 6. Very small grains of chromite (0·1 mm) were observed in the cumulate on Moskenesøy (GM 72).

Table 6. Electron microprobe analyses of ilmenite from mangeritic rocks, where ilmenite was the only magmatic oxide phase.

Feldspars

The plagioclase from the gabbro and anorthosite ranges in composition from An₆₉ to An₄₄ ( Fig. 7). Plagioclase compositions from the Moskenesøy cumulate could not be obtained because of extensive alteration to zoisite needles and secondary amphibole during later metamorphism. Plagioclase compositions in the anorthositic rocks range from An₅₇ to An₄₇ in the FBC and from An₄₈ to An₄₄ in the Eidsfjord anorthosite ( Fig. 7). According to Carstens, (1957), phenocrysts in the porphyritic border zone of the FBC near Napp have compositions of An₅₄Or₄, whereas the groundmass has An₄₅Or₉. There is no correlation between the normative plagioclase content (An + Ab + Or) in the anorthosites and the An content in the plagioclase ( Fig. 7b). Representative microprobe analyses of plagioclase grains and XRF analyses ofplagioclase separates are given in Table 7.

Table 7. Electron microprobe (EMP) and XRF analyses of plagioclase grains and mineral separates from anorthositic rocks on the Lofoten Islands.

Figure 7. (a) Plagioclase compositions from anorthositic rocks plotted within the An-Ab-Or triangle. The data include microprobe point analyses as well as XRF analyses of mineral separates. (b) Normative X_An and X_Or in anorthositic plagioclase plotted against the normative plagioclase content in anorthositic rocks from the FBC. These data are derived from whole-rock analyses.

The plagioclase analysed by electron microprobe is poorer in potassium (Or_0·8-2·1) than the calculated normative plagioclase composition (Or_2-9) from the same rock. Comparably high values are known from some other anorthosites [see Ashwal, (1993) and references therein]. We contend that this difference may reflect low-temperature exsolution of K-feldspar from the plagioclase as this was observed in sample GM 527, and also the inclusion of late K-rich interstitial liquid in the normative calculation.

The incorporation of interstitial liquid should be reflected in higher than normal Rb and Ba content in the analyses of the plagioclase separates. Comparison with data from the literature reveals that the Lofoten Rb and Ba contents (Rb: 12-21 ppm, one sample with 64 ppm; Ba: 310-930 ppm) are higher than would be expected for anorthositic plagioclase. For example, Ba in plagio- clase from the Egersund Ogna massif of Rogaland varies between 20 and 200 ppm (Duchesne & Demaiffe, 1978), and that from the Nain plutonic suite varies between 105 and 360 ppm (Gill & Murthy, 1970); Rb in plagioclase from Bjerkrem-Sogndal varies between 0·2 and 3·4 ppm (Duchesne, 1978), and that from Nain varies between 0·5 and 8·5 ppm (Gill & Murthy, 1970). Hence, the separate analyses appear not to represent the `true' composition of the anorthositic plagioclase but rather a mixture with minute grains of late, interstitial K-feldspar crystallized from late-stage intercumulus liquids.

Alkali feldspar

Ferrodiorite, jotunite and some relatively Mg-rich mangerites contain two strongly exsolved feldspars, whereas more evolved mangeritic rocks contain a single ternary alkali feldspar. Ternary feldspar compositions were calculated from whole-rock analyses assigning all K, Na and Al to the feldspar, because reintegration using image processing techniques gave unsatisfactory results (see below). The errors caused in the calculation of the ternary feldspar composition by excluding the Na and Al incorporation into clinopyroxene are small as a result of the large feldspar/augite ratio and the relatively low amounts of Na and Al in the clinopyroxene ( Table 3). The calculated ternary feldspars define a fractionation trend within the mangerite-charnockite series ( Fig. 8); they range from An₇Ab₅₃Or₄₀ in a Cpx-Opx mangerite to An_0·4Ab_53·6Or₄₆ in a Cpx-Ol-Qtz charnockite.

Figure 8. Normative feldspar compositions in mangeritic and charnockitic rocks in the An-Ab-Or triangle. The method of calculation is explained in the text. It should be noted that some of the points represent one ternary feldspar, some the mixture of two. The inset shows the boundary between the two- and the single-feldspar region. The solvus lines in the large triangle are according to Fuhrman & Lindsley, (1988).

Amphibole

Magmatic amphibole occurs in a few samples from Hamarøy. Typically, the amphibole is ferropargasitic hornblende containing small amounts of Cl (0·24-0·65 wt %, Table 8) that rims and replaces magmatic clinopyroxene and ilmenite, but coexists with orthopyroxene. The TiO₂ content of the magmatic amphiboles is relatively high (between 0·16 and 0·26 p.f.u., commonly >0·2 p.f.u., Table 8) in contrast to the Ti content of later metamorphic amphiboles (commonly <0·2 p.f.u.) which also coexist with ilmenite. Representative analyses of magmatic amphiboles are given in Table 8.

Table 8. Selected microprobe analyses of magmatic amphibole (ferropargasitic hornblende) from Hamarøy, Lofoten Islands.

PHASE PETROLOGY AND ESTIMATION OF INTRINSIC VARIABLES (P, T, f_O2, a_SiO2, f_HCl)

The mineral assemblages including pyroxenes, olivine and Fe-Ti oxides in the anorthositic and mangeritic rocks of the Lofoten Islands allow calculation of P, T, a_SiO2 and f_O2 using silicate-oxide (QUILF) equilibria (Frost & Lindsley, 1992; Lindsley & Frost, 1992; Andersen et al., 1993). Not only can the equilibrium conditions be estimated, the program can also be used to determine the extent to which the phases have changed composition during cooling. To do this, it is crucial to have knowledge of the original mineral assemblage, good estimates of the primary composition of at least one phase and an estimate of the relative volumes of the various ferromagnesian minerals in the rock. In most rocks, the primary phase of known composition was olivine and/or reintegrated clinopyroxene.

Pressure and temperature

The anorthositic rocks can be divided into three groups based on ferromagnesian silicate mineral assemblages, all of which contain magnetite and ilmenite. With increasing X_Fe the assemblages are: orthopyroxene, subcalcic augite and olivine (FBC); orthopyroxene and subcalcic augite (Eidsfjord); olivine and subcalcic augite (FBC). The various assemblages reflect differences in the intensive parameters during crystallization of these rocks.

Temperatures were calculated from pyroxene equilibria using the reintegrated mineral compositions of clinopyroxene and orthopyroxene with the QUILF program of Andersen et al., (1993). Because the pyroxene thermometer is slightly pressure dependent, we held pressure constant over a range of reasonable crustal conditions. The calculated temperatures within uncertainty for samples GM 265 and GM 139 from the FBC cover the range 1140-1185°C; cpx-opx assemblages from sample GM 452 gave temperatures of 1100-1140°C at pressures fixed in the range 4-9 kbar. The clinopyroxene-orthopyroxene pairs in ferrodiorite GM 451 were too exsolved to allow reintegration by any method.

There are three equilibria that are dependent on pressure, temperature and silica activity in rocks with the mineral assemblage olivine + orthopyroxene + clinopyroxene + plagioclase. These are

Reaction (3) is independent of reactions (1) and (2), but reactions (2) and (3) both depend on the Al₂O₃ content in Cpx because increasing pressure drives reactions (1) and (2) to the right. On the other hand, these equilibria are not affected by later Fe-Mg re-equilibration as equilibrium (3) is. For reactions (1) and (2), we used the GEOCALC software of Berman & Perkins, (1987) with the thermodynamic database of Berman, (1988). The activities of anorthite, albite, jadeite and Ca-tschermakite were calculated from the measured plagioclase and clinopyroxene compositions using the solution models of Fuhrman & Lindsley, (1988), Holland, (1990) and Wood, (1979), respectively. From reaction (3), a_SiO2 was calculated using the QUILF program of Andersen et al., (1993).

Pressure can be calculated by solving two of the three reactions (1)-(3) simultaneously. Hence, we have three independent ways of estimating the magmatic P-T conditions of the anorthositic rocks of the FBC. First, we simultaneously solved reactions (1) and (3) for various samples from the FBC ( Fig. 9a). In a second step, we simultaneously solved reactions (1) and (2) ( Fig. 9b) at constant silica activity for sample GM 333. This sample lacks olivine and orthopyroxene and therefore a silica activity could not be calculated from equilibrium (3). Instead, a reasonable range of values between 0·58 and 0·65 (see Fig. 9a) was chosen. Pressure changed by ~1·5 kbar within this range of silica activities. In a third step, we solved all three equilibria simultaneously for temperatures varied between 1100 and 1200°C, which is the range indicated by pyroxene thermometry. Figure 9c and d shows the resulting variation of pressure vs a_SiO2. The three equilibria give a reasonable intersection. The results reveal that the anorthositic rocks equilibrated over a range of pressures from ~9·5 ± 1·5 kbar in sample GM 255 and GM 265 to 7·5 ± 1·5 kbar in sample GM 333 and to 6 ± 1·5 kbar in sample GM 139 at temperatures in the range 1100-1200°C. The two samples GM 139 and GM 265 record distinctly different pressure conditions which are clearly outside of analytical uncertainty. Sample GM 333 falls in the range between sample GM 139 and sample GM 265. Figure 9a illustrates that the combination of equilibria (1) and (3) indicates an increase in silica activity from 0·58 to 0·65 with decreasing pressure, but this is neither supported nor refuted by the other calculations. The changes in pressure recorded in these samples are caused by the variation of Al₂O₃ content in clinopyroxene within and between samples (see Table 4). The continuous fractionation of plagioclase from An ₅₇ to An₄₇ within the FBC is superimposed on this pressure decrease.

Figure 9. (a) Pressure-temperature diagram where calculations of reaction (3) from QUILF (Andersen et al., 1993) are combined with the CaTs-An-Qtz reaction (1) to obtain estimates of the intrusion depth of the FBC. (b) Pressure-temperature diagram for sample GM 333 constructed by solving reactions (1) and (2) simultaneously with the GEOCALC software of Berman & Perkins, (1987). (c) and (d) Pressure vs silica activity diagrams for samples GM 255 and GM 265 (c) and GM 139 (d) calculated solving reactions (1), (2) and (3) simultaneously at temperatures fixed in the range 1100-1200°C. The numbers in parentheses correspond to the reaction numbers in the text. The methods are explained in detail in the text.

Only a maximum pressure could be obtained for the Eidsfjord anorthosite by introducing a fictive olivine in equilibrium with the observed ortho- and clinopyroxene in sample GM 452 using the QUILF program (Andersen et al., 1993). The composition of this olivine would be Fo₅₈ and the calculated maximum pressure and temperature would have been 7·3 kbar at 1135°C.

The uncertainty in the pressure and temperature calculations has three sources: (1) analytical uncertainty, (2) uncertainty in thermodynamic solution models and (3) uncertainty in the least-squares fit of the QUILF calculations. The uncertainty contributed by analytical error far overwhelms the other sources. Its effect was determined by using the whole variety of pyroxene and/or olivine compositions measured in one sample in the calculations which gave `extreme' values of pressure and temperature. All these factors together result in pressure uncertainties on the order of ±0·7 to ±1 kbar. The uncertainties become smaller by ~0·3 kbar when it is assumed that sample GM 265 crystallized at temperatures higher by ~50°C than GM 139 as is indicated by the equilibria in Fig. 9a. As a `worst case' approximation, the effect of analytical uncertainty in the Al₂O₃ measurements for clinopyroxene from sample GM 265 was evaluated. Deviation by 10 relative % (0·3-0·4 wt %) results in a pressure change of ~0·8 kbar and a temperature change of ~30°C; by 20% in a pressure change of 1·6 kbar and a temperature change of 60°C. As it is reasonable to assume that careful microprobe analyses do not result in errors of this size, we conclude that pressure differences larger than 1·5 kbar do not result from uncertainties in our calculations.

The pressure calculated for GM 265 based on the Al₂O₃ content of orthopyroxene (after Longhi et al., 1993) is considerably lower (~5 kbar in contrast to ~9 kbar) than the pressure derived from the equilibria (1)-(3). However, Al in orthopyroxene is also strongly dependent on temperature and the experiments of Longhi et al., (1993) were performed at temperatures higher than the highest estimates that we derived for the FBC. Because of the steep P-T dependence of Al solubility in orthopyroxene, a 50°C difference results in a pressure difference of ~3·5 kbar according to the Al-in-orthopyroxene thermometer of Aranovich & Berman, (1997).

Mangeritic rocks

Mangeritic and charnockitic rocks exhibit four different ferromagnesian silicate assemblages. With increasing FeO^tot/(FeO^tot + MgO) (= X_Fe in whole rock = X_Fe^rock) these are ( Fig. 10):

orthopyroxene + subcalcic augite (SWLM) (0·81 < X_Fe < 0·90, X_Fe ^rock of whole rock); orthopyroxene + subcalcic augite + olivine (RMI)(X_Fe^rock = 0·9-0·92); olivine + inverted pigeonite + subcalcic augite + quartz (SØM and RMI) (X_Fe^rock = 0·92-0·96); olivine + subcalcic augite (SØM and RMI) + quartz (X_Fe^rock > 0·96).

Figure 10. Fe-Mg silicate assemblages in mangeritic and charnockitic rocks and their relation to the X_Fe [FeO^tot/(FeO^tot + MgO)] ratio of the whole rock.

The assemblages plotted in Fig. 10 come from all the intrusions studied on the Lofoten Islands. The fact that they all show a similar trend of mineral assemblages as a function of X_Fe^rock is strong evidence that all the mangeritic rocks of Lofoten crystallized from melts of similar compositions under similar P-T conditions. Ormaasen, (1976), however, reported two different trends among the mangeritic rocks: a high X_Fe^rock trend and a low X _Fe^rock trend.

The most valuable mineral assemblage for determining intensive parameters is augite-pigeonite-olivine-quartz (GM 254 and GM 394) because we can deduce P and T from it, even though the Fe/Mg ratios of the ferromagnesian silicates may have changed during cooling (Frost & Lindsley, 1992). Furthermore, if there are two Fe-Ti oxides coexisting (even if they are highly exsolved), the oxygen fugacity at which the assemblage equilibrated can be calculated.

Based on the textures in GM 254 we infer the former presence of olivine in this sample in which no remaining primary olivine is present. However, we observed textures that closely resemble corona textures found in samples still containing fayalite. There, the fayalite is surrounded and, in a few samples, totally replaced by orthopyroxene and an outer rim of garnet. This orthopyroxene is distinctly more Fe rich than orthopyroxene exsolved from clinopyroxene within the same sample. In GM 254, pigeonite was also present, but it was so completely exsolved that quantitative reintegration was not possible ( Fig. 4). The only phase whose primary composition could be determined was augite ( Table 3), although we know pigeonite, olivine and quartz were present. This information is sufficient to calculate the equilibrium assemblage and the equilibrium pressure and temperature of crystallization. The results are 3·9 ± 0·3 kbar and 860 ± 20°C (see Fig. 11 for complete graphical representation of the measured and calculated mineral compositions). The uncertainties result from the variation in augite compositions that we determined using a defocused electron microprobe beam ( Fig. 11). As with the anorthositic rocks, these measurement uncertainties are believed to cause the largest errors in P-T estimations.

Figure 11. An enlarged portion of the pyroxene quadrilateral showing the method that was used to estimate magmatic conditions in sample GM 254. (See text for explanation.)

The silicates in sample GM 394 give an equilibrium temperature of 480°C ( Fig. 12). This reflects extensive subsolidus exchange of Fe and Mg. During this subsolidus exchange the olivine becomes more Fe rich and the clinopyroxene more magnesian. In rocks with only a small amount of olivine (e.g. GM 254) the effect of this exchange on pyroxene composition is small; in rocks with a higher modal amount of olivine, such as GM 394, re-equilibration was more extensive. To quantify this effect, we estimated clinopyroxene-olivine ratios. For this purpose, the wt % analyses were converted into moles, all K, Na, P and Al were assigned to feldspars and apatite, and all Ti to ilmenite. The appropriate amount of Ca for these minerals from the original Ca value was then subtracted, a new X_Fe value was calculated and the rest of the Ca was assigned to clinopyroxene. Finally, the olivine was computed from the remaining Mg that was not used for clinopyroxene. Disregarding magnetite gives a minimum estimation of the olivine content. The calculated clinopyroxene-olivine ratios in samples from the Hamarøy part of the Raftsund mangerite are about 1:1 with a little variation (see box in Fig. 12), whereas in sample GM 254 this ratio is about 2:1.

Figure 12. An enlarged portion of the pyroxene quadrilateral showing the method that was used to estimate magmatic conditions in sample GM 394. (See text for explanation.)

We were, however, able to `see through' the low-temperature Fe-Mg exchange in sample GM 394 by using the measured compositions of olivine and augite ( Tables 2 and 3) and the knowledge that pigeonite, albeit strongly exsolved, had been present. In detail, we proceeded as follows. We performed two sets of calculations using the QUILF program. First we fixed the measured clinopyroxene composition and calculated pressure, temperature and the compositions of olivine and pigeonite that would have been in equilibrium with it. Then we fixed the measured olivine composition and calculated pressure, temperature and the equilibrium compositions of the clinopyroxene and pigeonite that would have been in equilibrium with the olivine. These calculations resulted in the white triangles of Fig. 12. It is evident from Fig. 12 that, before diffusive re-equilibration, the original compositions plotted between the two white triangles. It can be inferred that the olivine became richer in Fe and the clinopyroxene became richer in Mg.

To correct for the Fe-Mg exchange on cooling and assuming closed system behaviour, we estimated the bulk ratio of olivine to augite in GM 394 from a whole-rock analysis as described above. This `modal composition' (with respect to augite and olivine ratio; box in Fig. 12) was then used as an anchor point around which we let the olivine-clinopyroxene tie-line (bold dashed line, Fig. 12) rotate until it had the same slope as the olivine-clinopyroxene tie-lines in the white triangles. This was done purely graphically. Figure 12 illustrates that two extremes could be thought of where the original olivine-augite tie-lines were lying. These extremes were the most augite-rich estimate and the most olivine-rich estimate. The P-T values and the pigeonite compositions for both tie-lines and for the average of both were calculated (grey triangles in Fig. 12). The results were again 3·9 ± 0·3 kbar at 860 ± 30°C.

A similar procedure was used to calculate the temperature of the jotunite GM 384. Here, the pressure was chosen as 4 kbar (assuming that the jotunite crystallized at the same crustal level as the mangeritic rocks) and the temperature was calculated to be 925-945°C. The temperature of sample GM 378 (cpx-opx-ol) was calculated to be 890 ± 20°C at a given pressure of 4 kbar. Here, no significant Fe-Mg exchange between clinopyroxene and olivine at subsolidus conditions was observed.

In another approach to estimate temperatures and the temperature decrease during fractionation, we estimated the minimum temperature of crystallization for rocks that contained only a single feldspar (GM 542 and GM 18). In this calculation we used the normative feldspar composition as calculated from the whole-rock analyses ( Fig. 13). The temperature defined by these feldspars is ~925°C using the graphical estimation of the feldspar solvus of Fuhrman & Lindsley, (1988) (see Fig. 8). Lower temperatures are then indicated for rocks crystallizing two feldspars instead of one. From petrographic observations we infer that the transition from one to two feldspars must have occurred at about the same temperature as the transition from the assemblage Cpx + Opx to Cpx + Opx + Ol. For the latter assemblage, we calculated 890 ± 20°C from pyroxenes and olivine in sample GM 378. The ternary feldspar in sample GM 254, finally, yields a minimum temperature of ~850°C (after Fuhrman & Lindsley, 1988; see Fig. 8), which fits well with the pyroxene thermometry for the same sample (860 ± 20°C). The temperatures derived from our technique of feldspar thermometry are close to those obtained from pyroxenes, even though they are more likely to reflect liquidus compositions whereas the pyroxene-olivine temperatures derived from reintegrated mineral compositions and QUILF are probably closer to solidus temperatures.

Figure 13. Normative feldspar composition of mangeritic rocks plotted against the X_Fe [FeO^tot/(FeO^tot + MgO)] ratio of the whole rock. The separation of rocks with one feldspar from those with two primary feldspars is shown as a dashed line. (See text for discussion and explanation.)

Magmatic amphibole replacing clinopyroxene is present in several samples of charnockite with the highest X_Fe. To test the interpretation that these amphiboles are magmatic, and to determine an additional independent temperature to further test our results, we used the graphical Ti-in-amphibole thermometer of Spear, (1981) for grains that coexist with ilmenite ( Fig. 14, Table 8). This thermometer is essentially independent of pressure and calculated temperatures are in the range 790-845°C, which is in good agreement with the temperatures derived from QUILF, taking into account that amphibole crystallization postdates that of clinopyroxene and should therefore be at somewhat lower temperatures.

Figure 14. Ti-in-amphibole thermometry after Spear, (1981). The Ti content in magmatic amphiboles from RMI charnockites is plotted against their Cl content and temperature. The experimental data of Spear, (1981) are also shown.

f_O2

The oxygen fugacity of crystallization could be estimated using QUILF in some anorthositic, one jotunitic and some mangeritic rocks ( Fig. 15). In rocks that contain both ilmenite and magnetite, it was also possible to recalculate the magmatic oxide compositions, which were not directly measurable because of low-temperature exchange and resetting during cooling. In rocks with only one oxide (ilmenite), Ti exchange did not occur and thus the measured ilmenite composition could be combined with our previous estimates of P and T to derive the oxygen fugacity.

Figure 15. [Delta]FMQ [= f_O2(sample) - f_O2(fayalite-magnetite-quartz buffer)] vs X_Fe (whole-rock) plot for selected examples of anorthositic, jotunitic, mangeritic and charnockitic rocks from the Lofoten Islands. X_Fe = FeO^tot/(FeO ^tot + MgO); the two trends within the mangeritic rock group should be noted. The box with the arrow indicates a minimum estimate.

The anorthositic rocks of the FBC (samples GM 265 and GM 139) equilibrated at conditions slightly above those of the FMQ buffer ( Fig. 15). Sample GM 265 equilibrated at an oxygen fugacity ~0·76 log units above FMQ at 9 kbar and 1180°C, and sample GM 139 at ~0·62 log units above FMQ at 4 kbar and 1100°C. For the Eidsfjord anorthosite, we estimated minimum oxygen fugacities by calculating a fictive olivine that would be in equilibrium with the actually observed minerals. At 4 kbar, f_O2 would have been ~0·23, and at 9 kbar ~0·02 log units above FMQ.

The jotunites, mangerites, and charnockites from Lofoten equilibrated at progressively lower oxygen fugacities. The jotunite GM 384 is calculated to have equilibrated at 0·6 log units below the FMQ buffer. There are two series of mangeritic rocks, one relatively high f_O2 with ilmenite and magnetite and one relatively low f_O2 with only ilmenite ( Fig. 15). These may correspond to the low-X_Fe^rock and high-X_Fe^rock series of Ormaasen, (1976). In the mangeritic rocks, two oxides occur only in GM 394. The calculated equilibrium compositions of the oxides were Mt₄₃Usp₅₅ and Ilm₉₂Hem₇, and f_O2 was ~0·6 log units below the FMQ buffer. In the other rocks, ilmenite occurred without magnetite and these rocks gave oxygen fugacities of 0·9-1·5 log units below the FMQ buffer.

a_SiO2

When plotted against X_Fe, the calculated silica activities in the Lofoten rock types increase steadily from values of ~0·58 in the anorthositic rocks to 1·0 in the mangeritic and charnockitic rocks that are quartz saturated ( Fig. 16), where a_SiO2 is unity for pure SiO₂ at P and T. In samples GM 262 and 452, minimum values of a_SiO2 are obtained by introducing a fictive coexisting olivine. For GM 452, this value is 0·64 at 7·3 kbar and QUILF calculations revealed that the composition of the fictive olivine would have been Fo₅₈. The values of a_SiO2 in both of the anorthosite complexes are similar, although the Eidsfjord anorthosite may have slightly higher values. In mangerite GM 262, olivine would have been Fo₃₃ and the minimum a_SiO2 was calculated to be 0·6 at 4 kbar. A mangerite and a jotunite without quartz, but with assemblages more suitable for a_SiO2 calculation give a_SiO2 values of 0·83 (jotunite, GM 384) and 0·87 (mangerite, GM 378), respectively.

Figure 16. a_SiO2 vs X_Fe (whole-rock) plot for selected examples of anorthositic, jotunitic, mangeritic and charnockitic rocks from the Lofoten Islands. X_Fe = FeO^tot/(FeO ^tot + MgO). Boxes with an arrow indicate minimum estimations.

f_H2O

Mangeritic magmas are generally considered to have crystallized at very low water fugacities (e.g. Kolker & Lindsley, 1989) because of their anhydrous mineral assemblages and CO₂-rich fluid inclusions (e.g. Frost & Touret, 1989) Several mangeritic rocks from Lofoten (GM 391, 393, 394, 397), however, contain magmatic amphibole that partially replaces subcalcic augite. Interestingly, orthopyroxene remains stable and clinopyroxene is still present in most cases. All of these samples are olivine bearing, and so it is possible, knowing pressure, temperature and a_SiO2 from QUILF or the presence of quartz, to estimate the water activity in these magmas using the equilibrium

The equilibrium constant can be expressed as follows:

(5)

log K = -[Delta]G°_r/2·303RT, where [Delta]G°_r is the Gibbs free energy of the reaction at P and T, R is the gas constant and T is the temperature in Kelvin. The water activity is calculated to be

(6)

Because the conditions of formation for the amphiboles are estimated to be 4 kbar and 800°C, these activities can be converted to fugacities via the relation

(7)

where f*_H2O is the fugacity of pure water at 4 kbar and 800°C. This value is 3356 bar (Helgeson & Kirkham, 1974).

[Delta]G°_r was calculated using the SUPCRT92 program of Johnson et al., (1992) and their thermodynamic database. We assumed that the rock last equilibrated at 800°C and 4 kbar. We used activities for Mg₂SiO₄ in olivine and CaMgSi₂O ₆ in clinopyroxene as calculated from QUILF (Andersen et al., 1993) and the solution model of Will & Powell, (1992) for amphibole. With these data we calculated water fugacities in the range of 14-170 bars where the amphibole still coexisted with augite and from 35 to 510 bars where it no longer coexisted with augite (see Table 9). These values are considerably smaller than the values derived from the fayalite-annite-quartz-K-feldspar equilibrium by Kolker & Lindsley, (1989), who estimated water fugacities between 1700 and 3100 bars at 750-800°C and 4·5 kbar for quartz syenites in the Laramie anorthosite complex in Wyoming, and should be considered as minimum values.

Table 9. Calculated values of a_H2O, f_H2O and f_HCl in samples from the RMI, Hamarøy, Lofoten.

f_HCl

The magmatic amphiboles from the RMI contain a small amount of Cl (0·28-0·63 wt %, Table 8, Fig. 14). Unfortunately, there does not exist a calibration of a Cl-OH exchange reaction of the type

for which reason we are dependent on the biotite equilibria investigated by Munoz & Swenson, (1981):

Although our rocks do not contain magmatic biotite, the composition of a fictive biotite in equilibrium with the magmatic amphibole can be calculated because coexisting metamorphic amphiboles and biotites in the mangeritic rocks from Lofoten show a relatively constant distribution of Cl between biotite and amphibole. The distribution coefficient is ~0·35-0·36 and in these Fe-rich rocks it is independent of the X_Fe in the minerals (Markl & Bucher, 1995; Markl et al., in press). We can then deduce ratios of f_H2O/f_HCl ( Table 9) using the expression given by Munoz & Swenson, (1981). If these are combined with f_H2O values calculated from the amphiboles and with the known pressure of ~4 kbar, we can derive the f_HCl during the late stages of crystallization, when the amphiboles formed. The values of f_HCl lie in the range 9-305 bars ( Table 9). Although these values depend on many assumptions (e.g. that the distribution coefficient is essentially independent of temperature in the range 700-800°C and that it is also independent of X_Fe in the minerals) we believe the calculated values of f_HCl give at least the order of magnitude of the values during crystallization. The f_H2O/f_HCl ratio never exceeds two in the Lofoten charnockites.

DISCUSSION AND CONCLUSIONS

Summary of the results

Mineral chemistry variations and estimation of intrinsic variables show that there is a series of trends from primitive gabbroic to highly evolved (Fe-enriched) charnockitic rocks on the Lofoten Islands that may indicate a close relationship among the parental magmas of the various types of rocks that are typically observed to occur together within Proterozoic anorthosite massifs. These trends include the following:

(1) progressive Fe enrichment in olivine and pyroxenes;
(2) feldspar evolution from calcic plagioclase via two-feldspar assemblages to strongly ternary and finally to nearly Ca-free alkali feldspar;
(3) temperature estimations show a decrease from those near the solidus of primitive basalt in the Moskenesøy cumulates (probably on the order of 1250-1300°C) and from 1185°C in the anorthosites to ~800°C in the latest charnockites;
(4) pressure estimations indicate initial crystallization of the anorthosite at the base of the crust (~9 kbar) and later ascent to pressures of ~4 kbar, where the mangerites and charnockites also intruded;
(5) silica activity shows increase from values ~0·58 in the anorthosites to quartz saturation in the charnockites;
(6) f_O2 ranged from 0·7-0·9 log units above FMQ in the anorthosites to 0·6-1·5 log units below FMQ in the charnockites;
(7) water fugacities were very low during crystallization of both the anorthositic and the mangeritic rocks, and allow crystallization of magmatic amphibole only in the most evolved charnockitic rocks.

Comparison with previous studies on the Lofoten Islands

Our results do not agree with previous studies (RMI: Krogh, 1977; HMI: Ormaasen, 1977), which estimated that the Lofoten mangerites equilibrated at 1000-1050°C at 9-12 kbar. This is significant in understanding the geology of northern Norway because Griffin et al., (1978) used these values to conclude that the Lofoten terrane is an unusual deep section through the crust. One cannot argue that this discrepancy is caused by the fact that our samples represent higher portions (and therefore lower pressures) of the same intrusions, because, as indicated by the HMI, the mangerite bodies are not thicker than ~3·5-4 km and the pressure within a single magma chamber should not vary by >1 kbar.

Ormaasen, (1977) calculated magmatic pressures from the same mineral assemblage that we used, namely olivine-orthopyroxene-quartz. We interpret the difference in values as a consequence of the fact that Ormaasen, (1977) used outdated evaluations of these equilibria and that mineral compositions were reset during later metamorphism. The pressure estimations of Krogh, (1977) for the basement rocks east of the RMI (750-765°C at ~9 kbar) are based on two-pyroxene, Al-in-pyroxene and pyroxene-garnet thermometry. These assemblages probably record the later high-pressure metamorphism that affected the Lofoten Islands and do not reflect the emplacement level of the mangerites.

The magmatic temperature estimations of Ormaasen, (1977) and Krogh, (1977) are based mainly on Fe-Ti oxide thermometry using the graphical calibration of Buddington & Lindsley, (1964). It is well known, however, that oxide thermometry cannot retrieve magmatic temperatures in plutonic rocks (Duchesne, 1972; Frost et al., 1988). The temperature conditions calculated by Griffin et al., (1974) and Malm, (1976) are based on pyroxene equilibria, but in both studies, reintegration techniques and textures such as the extent of exsolution were not described in detail. Calculations with QUILF using the analyses of Griffin et al., (1974) or Malm, (1976) gave temperatures of the order of 500°C and pressures between 9 and 12 kbar.

The finding that the mangerites intruded at 4 kbar throughout the Lofoten Islands is especially important, because gravity and seismic data (Svela, 1971; Mjelde et al., 1993) indicate that the crust is only ~18-19 km thick under the southwestern islands, whereas it is 25-28 km thick below the middle and eastern islands. This crustal structure is also reflected in a large positive gravity anomaly under Moskenesøy (Svela, 1971). Despite the fact that the crustal thickness below the Lofoten Islands decreases from east to west, the crustal level observed at the surface is the same throughout the Lofoten Islands. Thus, the mantle upwelling that has occurred under Moskenesøy since the Cretaceous (Mjelde et al., 1993) did not result in enhanced uplift and erosion of the western islands, but must evidently be accommodated by plastic flow and delamination of the lower parts of the crust above the upwelling mantle.

Comparison with other Proterozoic anorthosite complexes

This is the first study to present direct constraints for the polybaric crystallization of anorthosites, which was suggested by Emslie, (1978) and many workers subsequently. Earlier studies used barometry in the country rocks (e.g. Berg, 1977) or in accompanying granitoid intrusives (e.g. Fuhrman et al., 1988) to estimate the crystallization pressure of anorthosites. Such studies derived information only about the final level of emplacement, but not about a crystallization history of the anorthosites themselves.

The pressures and temperatures of final emplacement of both the anorthosites and the related mangeritic rocks derived in this study are consistent with estimated conditions of intrusion of anorthosite-mangerite- charnockite-granite (AMCG) complexes from all over the world (e.g. Berg, 1977; Duchesne, 1984; Fuhrman et al., 1988; Kolker & Lindsley, 1989; Ashwal, 1993; Frost et al., 1993). Calculated mangerite crystallization temperatures in this study are somewhat lower, but in the same range as those reported by Fuhrman et al., (1988) for the Sybille Monzosyenite in the Laramie Anorthosite Complex, and both temperatures and pressures agree perfectly with those of Frost & Bucher, (1993) from Dronning Maud Land, Antarctica, where large amounts of mangeritic to charnockitic rocks are spatially associated with two Proterozoic anorthosite complexes (Kämpf & Stackebrandt, 1985).

Polybaric crystallization of anorthosites: interpretations

Since 1978, polybaric crystallization has been the main tenet of nearly all studies on Proterozoic anorthosites (e.g. Emslie, 1978; Morse, 1982; Duchesne, 1984; Wiebe, 1992; Ashwal, 1993; Longhi et al., 1993). The present study is the first to use phase equilibria in natural samples from anorthosites to confirm this and therefore has-together with the experimental work of Fram & Longhi, (1992) and Longhi et al., (1993)-considerable bearing on the evaluation of current ideas about the formation of Proterozoic anorthosite complexes.

The assumption of a polybaric crystallization history for anorthosites results from various observations:

(1) the occurrence of high-Al ortho- or clinopyroxene megacrysts with 4-11·9 wt % Al₂O₃ (Emslie, 1975; Longhi et al., 1993), which are believed to provide evidence of earlier crystallization at greater depths (15 ± 5 kbar), although Morse, (1975) and Dymek & Gromet, (1984) interpreted high-Al orthopyroxene megacrysts as having formed because of rapid metastable crystallization at the final level of emplacement of the anorthosite and hence at relatively low pressures;
(2) the intermediate plagioclase that crystallizes from basaltic magmas only under high pressures (Fram & Longhi, 1992);
(3) the lack of ultramafic cumulates and of gravity anomalies that would be produced by ultramafic cumulates; these cumulates should occur as complementary rocks to the anorthosites when these crystallize from a basaltic magma; as they are not observed they are believed to have remained at the base of the crust.

All these observations can be best explained when crystallization starts at deep levels and the anorthosites intrude as crystal-rich mushes (Emslie, 1978; Fram & Longhi, 1992; Wiebe, 1992).

The pressures of crystallization determined in this study indicate a polybaric emplacement history for the Lofoten anorthosites. We note that the decompression recorded in the pyroxenes is the minimum likely decompression because pyroxenes and olivine crystallized texturally late and the pressure change recorded is only that which occurred after pyroxene crystallization began. The associated mangeritic rocks crystallized at pressures similar to the lowest pressures found in the anorthosites. As the plagioclase composition in the Lofoten anorthosites changes only little during this uprise [because of the contrasting effects of decreasing pressure and continuing fractionation (Fram & Longhi, 1992)], the effect of decreasing pressure is recorded mainly by the interstitial mafic silicates, most importantly by the Al₂O₃ content in clinopyroxene. These changes in the CaTs content of the clinopyroxene reflect a pressure decrease from >9 to ~5-6 kbar and a temperature decrease from 1185 to 1150°C during crystallization when the phase equilibria among the mafic silicates are combined with plagioclase equilibria. Continued fractionation during uprise is recorded by olivine and plagioclase, which change their composition from An₅₇ to An₄₇ and from Fo₇₂ to Fo ₆₆, respectively. This is also supported by the occurrence of more calcic plagioclase phenocrysts within a more sodic matrix in the porphyritic border zone of the FBC (Carstens, 1957). Because some of the rocks show a range in pyroxene composition, this means that crystallization proceeded as the magma rose to mid-crustal levels. The large variations between different samples could suggest that more solidified parts were entrained as `rafts' in more liquid-rich parts during ascent and therefore could intrude to higher levels. This fits with the theory that some anorthosites were emplaced as crystal-rich mushes (Emslie, 1978; Morse, 1982; Fram & Longhi, 1992; Ashwal, 1993; Mitchell et al., 1995, , 1996; Scoates & Frost, 1996).

The present paper provides extensive phase petrological data and interpretations that can be taken as support for the two-stage model of anorthosite genesis with early, deep crystallization and later emplacement at shallower crustal levels, as well as for the geochemical evolution of mangeritic and charnockitic rocks associated with anorthosites. The techniques put forward and applied in this study may be of use in other anorthosite complexes or layered intrusions, and may possibly lead to a much better understanding of phenomena such as loci of crystallization, storage of magma and emplacement of intrusions.

ACKNOWLEDGEMENTS

This work greatly benefited from very careful reviews by J.-C. Duchesne, D. Lindsley, J. Scoates and J. Schumacher. We are very grateful for their suggestions and their criticism. We are also grateful to K. Fesenmeier, H. Schlegel, N. Kindler and Dr H. Müller-Sigmund, Universität Freiburg, for preparation of the thin sections and for their help during preparation and the analytical procedure. A. Kronz and B. Schulz-Dobrick, Universität Mainz, assisted during some of the microprobe measurements. E. Tveten, NGU Trondheim, gave very helpful comments for the field work on Langøy and Hamarøy. G. M. thanks A. Wamsler, Freiburg, for being a great field companion during two summers. This work was funded by the Deutsche Forschungsgemeinschaft with Grants Bu 843/3-1, 3-2 and 3-3.

REFERENCES

Andersen, D. J., Lindsley, D. H. & Davidson, P. M. (1993). QUILF: a PASCAL program to assess equilibria among Fe-Mg-Mn-Ti oxides, pyroxenes, olivine, and quartz. Computers and Geosciences 19, 1333-1350.

Aranovich, L. Y. & Berman, R. G. (1997). A new garnet-orthopyroxene thermometer based on reversed Al₂O₃ solubility in FeO-Al₂O₃-SiO₂ orthopyroxene. American Mineralogist 82, 345-353.

Ashwal, L. D. (1993). Anorthosites. Berlin: Springer-Verlag.

Berg, J. H. (1977). Regional geobarometry in the contact aureoles of the anorthositic Nain Complex, Labrador. Journal of Petrology 18, 399-430.

Berman, R. G. (1988). Internally-consistent thermodynamic data for minerals in the system Na₂O-K₂O-CaO-MgO-FeO-Fe₂O₃- Al₂O₃-SiO₂-TiO₂-H₂O-CO₂. Journal of Petrology 29, 445-522.

Berman, R. G. & Perkins, E. H. (1987). GEOCALC: software for calculation and display of pressure-temperature-composition phase diagrams. American Mineralogist 72, 861-862.

Buddington, A. F. & Lindsley, D. H. (1964). Iron-titanium oxide minerals and synthetic equivalents. Journal of Petrology 5, 310-351.

Carstens, H. (1957). Investigations of the titaniferous iron ore deposits. Part II: Gabbro, anorthosite, and titaniferous iron ore on Flakstadøy in Lofoten, Northern Norway. Kongelige Norske Videnskabers Selskab Skrifter 4, 2-24.

Duchesne, J. C. (1972). Iron-titanium oxide minerals in the Bjerkrem-Sogndal massif, south-western Norway. Journal of Petrology 13, 57-81.

Duchesne, J. C. (1978). Quantitative modeling of Sr, Ca, Rb, and K in the Bjerkrem-Sogndal layered lopolith (S.W. Norway). Contributions to Mineralogy and Petrology 66, 175-184.

Duchesne, J. C. (1984). Massif anorthosites: another partisan review. In: Brown, W. L. (ed.) Feldspars and Feldspathoids. Boston, MA: D. Reidel, pp. 411-433.

Duchesne, J. C. (1990). Origin and evolution of monzonorites related to anorthosites. Schweizerische Mineralogische und Petrographische Mitteilungen 70, 189-199.

Duchesne, J. C. & Demaiffe, D. (1978). Trace elements and anorthosite genesis. Earth and Planetary Science Letters 38, 249-272.

Duchesne, J. C., Wilmart, E., Demaiffe, D. & Hertogen, J. (1989). Monzonorites from Rogaland (southwest Norway): a series of rocks coeval but not comagmatic with anorthosites. Precambrian Research 45, 111-128.

Dymek, R. F. & Gromet, L. P. (1984). Nature and origin of orthopyroxene megacrysts from the St-Urbain anorthosite massif, Quebec. Canadian Mineralogist 22, 297-326.

Emslie, R. F. (1975). Pyroxene megacrysts from anorthositic rocks: new clues to the source and evolution of the parent magmas. Canadian Mineralogist 13, 138-145.

Emslie, R. F. (1978). Anorthosite massifs, rapakivi granites, and late Proterozoic rifting of North America. Precambrian Research 7, 61-98.

Fram, M. S. & Longhi, J. (1992). Phase equilibria of dikes associated with Proterozoic anorthosite complexes. American Mineralogist 77, 605-616.

Frost, B. R. & Bucher, K. (1993). Charnockites of the Thor Range, Queen Maud Land, Antarctica: the granulite uncertainty principle exemplified. Geological Society of America, Abstracts with Programs 25, 448.

Frost, B. R. & Lindsley, D. H. (1992). Equilibria among Fe-Ti oxides, pyroxenes, olivine, and quartz: Part II. Application. American Mineralogist 77, 1004-1020.

Frost, B. R. & Touret, J. L. R. (1989). Magmatic CO₂ and saline melts from the Sybille Monzosyenite, Laramie Anorthosite Complex, Wyoming. Contributions to Mineralogy and Petrology 103, 178-186.

Frost, B. R., Lindsley, D. H. & Andersen, D. J. (1988). Fe-Ti oxide-silicate equilibria: assemblages with fayalitic olivine. American Mineralogist 73, 727-740.

Frost, B. R., Frost, C. D., Lindsley, D. H., Scoates, J. S. & Mitchell, J. N. (1993). The Laramie Anorthosite Complex and Sherman batholith: geology, evolution, and theories of origin. Geological Survey of Wyoming Memoir 5, 118-161.

Fuhrman, M. & Lindsley, D. (1988). Ternary-feldspar modeling and thermometry. American Mineralogist 73, 201-215.

Fuhrman, M. L., Frost, B. R. & Lindsley, D. H. (1988). Crystallization conditions of the Sybille monzosyenite, Laramie anorthosite complex, Wyoming. Journal of Petrology 29, 699-729.

Gill, J. B. & Murthy, V. R. (1970). Distribution of K, Rb, Sr, and Ba in Nain anorthosite plagioclase. Geochimica et Cosmochimica Acta 34, 401-408.

Griffin, W. L., Heier, K. S., Taylor, P. N. & Weigand, P. W. (1974). General geology, age and chemistry of the Raftsund Mangerite Intrusion, Lofoten-Vesterålen. Norges Geologiske Undersøkelse 312, 1-30.

Griffin, W. L., Taylor, P. N., Hakkinen, J. W., Heier, K. S., Iden, I. K., Krogh, E. J., Malm, O., Olsen, K. I., Ormaasen, D. E. & Tveten, E. (1978). Archean and Proterozoic crustal evolution in Lofoten-Vesterålen, N Norway. Journal of the Geological Society, London 135, 629-647.

Hames, W. E. & Andresen, A. (1996). The timing of Paleozoic orogeny and extension in the continental shelf of northcentral Norway as indicated by laser ⁴⁰Ar/³⁹Ar muscovite dating. Geology 24, 1005-1008.

Heier, K. S. (1960). Petrology and geochemistry of high-grade metamorphic and igneous rocks on Langøy, northern Norway. Norges Geologiske Undersøkelse 207, 1-246.

Heier, K. S. & Compston, W. (1969). Interpretation of Rb-Sr age patterns in high grade metamorphic rocks, north Norway. Norsk Geologisk Tidsskrift 49, 257-283.

Helgeson, H. C. & Kirkham, D. H. (1974). Theoretical prediction of the thermodynamic behaviour of aqueous electrolytes at high pressures and temperatures: I. Summary of the thermodynamical/electrostatic properties of the solvent. American Journal of Science 274, 1089-1198.

Holland, T. J. B. (1990). Activities of components in omphacitic solid solutions. Contributions to Mineralogy and Petrology 105, 446-453.

Johnson, J. W., Oelkers, E. H. & Helgeson, H. C. (1992). SUPCRT92: a software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bars and 0 to 1000°C. Computers and Geosciences 18, 899-947.

Kämpf, H. & Stackebrandt, W. (1985). Geologic investigations in the Eliseev Anorthosite Massif, central Dronning Maud Land, East Antarctica. Zeitschrift für Geologische Wissenschaften 13, 321-333.

Kolker, A. & Lindsley, D. H. (1989). Geochemical evolution of the Maloin Ranch pluton, Laramie anorthosite complex, Wyoming: petrology and mixing relations. American Mineralogist 74, 307-324.

Krogh, E. J. (1977). Origin and metamorphism of iron formations and associated rocks, Lofoten-Vesterålen, N. Norway. I. The Vestpolltind Fe-Mn deposit. Lithos 10, 243-255.

Lindsley, D. H. (1983). Pyroxene thermometry. American Mineralogist 68, 477-493.

Lindsley, D. H. & Frost, B. R. (1992). Equilibria among Fe-Ti oxides, pyroxenes, olivine, and quartz: Part I. Theory. American Mineralogist 77, 987-1003.

Longhi, J., Fram, M. S., Vander Auwera, J. & Montieth, J. N. (1993). Pressure effects, kinetics, and rheology of anorthositic and related magmas. American Mineralogist 78, 1016-1030.

Malm, O. (1976). De mangeritiske bergarter med tilstøtende orthogneiser i Hamarøy, Nordland. Generell geologi, petrografi, geokjemi og petrologi. Cand. Real. Thesis, University of Oslo, 230 pp.

Malm, O. & Ormaasen, D. E. (1978). Mangerite-charnockite intrusives in the Lofoten-Vesterålen area, north Norway: petrography, chemistry and petrology. Norges geologiske Undersøkelse 338, 83-114.

Markl, G. & Bucher, K. (1995). Cl-bearing hydrosilicates in granulite-facies rocks from Lofoten, Norway. European Journal of Mineralogy, Beiheft 1 7, 162.

Markl, G. & Bucher, K. (1997). Proterozoic eclogites from the Lofoten Islands, N. Norway. Lithos 42, 15-35.

Markl, G., Ferry, J. & Bucher, K. (1998). Formation of saline brines and desiccation of the lower crust by retrograde hydration reactions in granulites from the Lofoten Islands, Norway. American Journal of Science 298, in press.

Misra, S. N. & Griffin, W. L. (1972). Geochemistry and metamorphism of dolerite dikes from Austvågøy in Lofoten. Norsk Geologisk Tidsskrift 52, 409-425.

Mitchell, J. N., Scoates, J. S. & Frost, C. D. (1995). High-Al gabbros in the Laramie Anorthosite Complex, Wyoming: inplications for the composition of melts parental to Proterozoic anorthosite. Contributions to Mineralogy and Petrology 119, 166-180.

Mitchell, J. N., Scoates, J. S., Frost, C. D. & Kolker, A. (1996). The geochemical evolution of anorthosite residual magmas in the Laramie Anorthosite Complex, Wyoming. Journal of Petrology 37, 637-660.

Mjelde, R., Sellevoll, M. A., Shimamura, H., Iwasaki, T. & Kanazawa, T. (1993). Crustal structure beneath Lofoten, N. Norway, from vertical incidence and wide-angle seismic data. Geophysical Journal International 114, 116-126.

Morse, S. A. (1975). Plagioclase lamellae in hypersthene, Tikkoatokhakh Bay, Labrador. Earth and Planetary Science Letters 26, 331-336.

Morse, S. A. (1982). A partisan review of Proterozoic anorthosites. American Mineralogist 67, 1087-1100.

Munoz, J. L. & Swenson, A. (1981). Chloride-hydroxyl exchange in biotite and estimation of relative HCl/HF activities in hydrothermal fluids. Economic Geology 76, 2212-2221.

Olsen, K. I. (1978). Metamorphic petrology and fluid-inclusion studies of granulites and amphibolite-facies gneisses on Langøy and W Hinnøy, Vesterålen, N Norway. Cand. Real. Thesis, University of Oslo, 214 pp.

Ormaasen, D. E. (1976). Geochemistry and petrology of the mangeritic rocks in Lofoten-Vesterålen. Cand. Real. Thesis, University of Oslo, 197 pp.

Ormaasen, D. E. (1977). Petrology of the Hopen mangerite-charnockite intrusion, Lofoten, north Norway. Lithos 10, 291-310.

Owens, B. E., Rockow, M. W. & Dymek, R. F. (1993). Jotunites from the Grenville Province, Quebec: petrological characteristics and implications for massif anorthosite petrogenesis. Lithos 30, 57-80.

Philpotts, A. R. (1981). A model for the generation of massif-type anorthosites. Canadian Mineralogist 19, 233-253.

Pouchou, J. L. & Pichoir, F. (1984). A new model for quantitative analysis. I. Application to the analysis of homogeneous samples. La Recherche Aérospatiale 3, 13-38.

Romey, W. D. (1971). Basic igneous complex, mangerite, and high grade gneisses of Flakstadøy, Lofoten, northern Norway: I. Field relations and speculations on origin. Norsk Geologisk Tidsskrift 51, 33-61.

Scoates, J. S. & Frost, C. D. (1996). A strontium and neodymium isotopic investigation of the Laramie anorthosites, Wyoming, USA: implications for magma chamber processes and the evolution of magma conduits in Proterozoic anorthosites. Geochimica et Cosmochimica Acta 60, 95-107.

Spear, F. S. (1981). An experimental study of hornblende stability and compositional variability in amphibolite. American Journal of Science 281, 697-734.

Svela, P. T. (1971). Gravimetriske undersøkelser av Lofoten-Vesterålen området. Cand. Real. Thesis, University of Bergen, 131 pp.

Vogt, T. (1909). Om eruptivbergartene paa Langøen i Vesteraalen. Norges Geologiske Undersøkelse 53, 1-39.

Wade, S. J. R. (1985). Radiogenic isotope studies of crust-forming processes in the Lofoten-Vesterålen Province of north Norway. Ph.D. Thesis, University of Oxford, 285 pp.

Wiebe, R. A. (1988). Structural and magmatic evolution of a magma chamber: the Newark Island Layered Intrusion, Nain, Labrador. Contributions to Mineralogy and Petrology 29, 383-411.

Wiebe, R. A. (1992). Proterozoic anorthosite complexes. In: Condie, K. C. (ed.) Proterozoic Crustal Evolution. Amsterdam: Elsevier, pp. 251-261.

Will, T. M. & Powell, R. (1992). Activity-composition relationships in multicomponent amphiboles: an application of Darken's quadratic formalism. American Mineralogist 77, 954-966.

Wood, B. J. (1979). Activity-composition relationships in Ca(Mg,Fe)Si₂O₆-CaAl₂SiO₆ clinopyroxene solid solution. American Journal of Science 279, 854-875.

---

*Corresponding author. Telephone: +49 761 203 6414. Fax: +49 761 203 6407. e-mail: markl@ruf.uni-freiburg.de

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals World Wide Web service.
Comments and feedback: www-admin@oup.co.uk
Last modification: July 1998
Copyright© Oxford University Press, 1998.