N G F A bstracts and Proceedings, N o. 2, 2003
57
1)
1) Geological Museum, University of Oslo, P.O. Box 1172 Blindern, –0318 Oslo, Norway.
(t.v.segalstad@nhm.uio.no & gunnar.raade@nhm.uio.no)
The chemical element scandium was named after Scandinavia by its discoverer Lars Fredrik Nilson in 1879. The reason for the naming was that he had found the element in the minerals euxenite and gadolinite, at that time only known from Scandinavian occurrences. For some reason Scandinavia, and in particular Norway, has been blessed with minerals containing this quite rare element. To be exact: scandium is not that rare, but because the element easily substitutes for the major elements ferric iron and aluminium in rock-forming ferromagnesian minerals, separate mineral species with scandium as a major element will be rare.
The main purpose of this contribution is to give an introduction to the geological background for known occurrences of scandium and scandium minerals in southern Norway. It must be stressed that the following text only presents a small part of the enormous material that has been published by many different authors from the field-trip localities, and that the present text by no means is meant to be an exhaustive account.
The Proterozoic (= younger part of Precambrian) gneisses of southern Norway are part of the Fennoscandian
Shield. The gneisses have undergone polyphase folding and metamorphism. The main deformation occurred in the Svecofennian (ca. 1,600-1,700 Ma) and Sveconorwegian (ca. 1,000-1,200 Ma) periods. The gneisses in this region form a series of NNW-SSE striking provinces, cut by the overall N-S trend of the Carbo-Permian (ca.
300-230 Ma) Oslo Rift and graben. The rocks have been extensively eroded by glaciers during the Quaternary period.
The central part of Oslo is situated on Cambro-Silurian sedimentary rocks, subsided approximately 1000 m in the O slo G raben. The subsidence in the southern Oslo Fjord has been estimated to some 3 km. The Carbo-
Permian igneous rocks of the Oslo Region were early known. The latitic lava and dike rock “rhomb porphyry”
(RP), with its typical rhomb-formed light-coloured feldspar phenocrysts occurring in a darker fine-grained groundmass, was first described from Oslo by the German geologist Leopold von Buch (1810). The rocks are rich in alkalis, and they were named from the places they were first found, for instance nordmarkite, lardalite, hedrumite, akerite, ekerite, larvikite.
During the Carbo-Permian period, southeastern Norway suffered an intracontinental rifting episode. The evolution of the Oslo Rift is described as follows: the activity started with sedimentary filling of an initial Carbo-
Permian trough, followed first by the initial basaltic volcanism (Segalstad 1978), then graben subsidence with fissure eruptions of the RP latite magma. This stage was followed by the subsidence of central volcanoes and cauldrons. The final stage was the emplacement of gabbro necks and major batholiths of syenite and granite
(Ramberg & Larsen 1978).
The chemical composition of the lavas ranges from silica-undersaturated alkaline basaltic rocks to tholeiite; intermediate RP to silica-oversaturated rhyolite and rhyodacite. The plutonic association embraces foyaite
(lardalite and hedrumite), monzonite (larvikite) with associated nepheline-syenite pegmatites (with rare

58
N G F A bstracts and Proceedings, N o. 2, 2003 minerals), syenite (nordmarkite etc.), alkali granite (ekerite), and biotite granite. Most of the dikes in the Oslo
Rift consist of diabase, some of syenite, and RP. Larvikite and RP are of similar intermediate chemistry, and are supposed to be the intrusive and extrusive representatives, respectively, of the same group of melts.
The Vestfold basaltic lavas have remarkably constant Sc contents near 30 ppm, except for the plagioclaseporphyritic lavas with a Sc content near 9 ppm, and some clinopyroxene-olivine porphyritic lavas with a Sc content near 44 ppm (Neumann et al. 1990). RP lavas and dikes show Sc contents from 8 to 15 ppm (Sundvoll
1994), while larvikite shows Sc contents of 5 to 17 ppm (Neumann 1980). Pyroxenes in pyroxenites have 100 ppm Sc, in kjelsåsite 70 ppm, in larvikite 150 ppm, in lardalite 100 ppm; biotites in pyroxenite 7 ppm, in larvikite
5 ppm, in lardalite < 3 ppm, in nordmarkite and granites 10 ppm Sc (Oftedal 1943).
mineral with REE + Th in a metasomatized anatase-albitite dike at Grorud, Oslo (Segalstad 1984). It is important to note that no scandium minerals have been reported from nepheline syenite pegmatites. In gadolinite-(Ce), chevkinite, and perrierite from the rare-mineral-rich nepheline-syenite pegmatites in the southwestern Oslo
Region, Sc was not detected by the electron microprobe (Segalstad & Larsen 1978a, 1978b). For Sc in larvikiteassociated minerals, see below under “The Pegmatites of the Southern Larvikite Area”.
The lower Paleozoic shales and limestones of the Oslo Region have 1.9 to 24.2 ppm (average 10.7 ppm) Sc. The shales contain generally near 15 ppm Sc, while limestone generally contain less than 10 ppm Sc (Dypvik 1977).
Radiogenic isotope studies of the igneous rocks give values typical of a mantle source for the magmas (Heier
& Compston 1969, Jacobsen & Wasserburg 1978, Anthony et al. 1989, Sundvoll & Larsen 1990). From petrographic evidence, Barth (1944) launched a petrogenetic model in which the Oslo igneous rock series evolved through crystal fractionation from a basaltic mother magma (Fig. 1). Further work showed that the mother magmas evolved through progressive melting of a rising mantle plume (Segalstad 1978, Neumann et al.
2002). However, some of the magmas have been modified through assimilation of crustal rocks (Jacobsen &
Wasserburg 1978, Segalstad & Ohmoto 1980). A dense gabbroic “pillow” is seen as a regional positive gravimetric anomaly 20-30 km underneath the present surface of the Oslo Rift (Ramberg 1976).
Fig. 1. Barth’s “family tree”, based on range in modal minerals, showing how the general development of the Oslo Region plutonic rocks could be explained from fractional crystallization from a monzonite (larvikite) magma, whose precursor was a mantle-derived basaltic magma. NE-PEGM. = nepheline-syenite pegmatite. Alk. amph. = alkali amphiboles. Redrawn from Barth (1944).

N G F A bstracts and Proceedings, N o. 2, 2003
59
Numerous metallic and non-metallic mineral deposits appear to be genetically associated with the Carbo-Permian rifting episode, and occur in both Permian, lower Paleozoic, and Precambrian host rocks. Just a few of these mineral deposits have been of any large economic importance. Several chemical elements in a wide variety of combinations and deposits are present. The metallic deposits include sulphide deposits with mainly Zn, Pb, Cu,
M o, and Bi; oxide deposits of Fe, Mn, Ti, and W; and deposits of native silver with Co-Ni arsenides. The types of deposits include magmatic segregation deposits, intramagmatic and perimagmatic vein deposits, porphyry molybdenum deposits, and contact metasomatic (skarn) deposits (Bugge 1978, Ihlen & Vokes 1978). Stable isotope work has shown that both magmatic and meteoric/crustal waters were involved in the making of the hydrothermally formed mineral deposits of the Oslo Rift (Segalstad & Ohmoto 1986, Segalstad & Telstø 2002).
The Kongsberg silver district (Bugge 1917, Ihlen & Nordrum 1986) is located close to the western margin of the Oslo Rift system. The native silver-bearing calcite veins are known to occur where the fracture system intersects "fahlbands", i.e., schistose zones in the Precambrian gneisses rich in iron sulphides (mainly pyrrhotite).
Ages of the hydrothermal veins have been given as 265 ± 3 Ma based on K-Ar isotope analysis of wall-rock alteration phyllosilicates (Ineson et al. 1975). The main ore-forming stages carry in mineralizing order mainly: quartz, pyrite, calcite (in five main generations), “coalblende” (a carboniferous substance), fluorite, galena, sphalerite, chalcopyrite, sulphosalts of silver, argentite, silver, and pyrrhotite (Neumann 1944). No scandium values have been published from here.
The calcite vein-fracture set is by Segalstad (1996) regarded as tension gashes which opened up as a response to shear movement along mylonite zones (locally called "rotten veins") which have their slip pointing south, possibly as a response to the Kongsberg block being tilted when it bent into the Oslo Graben in Permian times.
Fluid-inclusion temperatures of formation range between 200 and 300 E C (assuming hydrostatic pressure of 350 bar at 3.5 km stratigraphic depth; Segalstad et al., 1986). Salinities range from 0 to 35 equivalent wt.% NaCl.
The inclusion data of Johansen (1985) reveal at least three different heating and cooling cycles with associated salinity cycling. Native silver has generally been deposited during the heating part of the cycles from 250 to
300 E C associated with a decrease in salinity from about 27-22 to about 20-15 eq.wt.% NaCl (absolute values are different for different mines and different cycles). Mineralizing fluids which reached temperatures less than
250 E C and/or reached salinity maxima less than 22 eq.wt.% NaCl have not been found to deposit native silver.
contemporaneous Permian sea water, seeping down the veins, or from reworked Permian Oslo Rift evaporites.
“coalblende” points to Oslo Region shales as possible sources for the major part of the carbon in this substance shows that the mineralizing fluids must have been "crustal water", i.e. mainly meteoric water which has exchanged oxygen isotopes with the crustal rocks (Segalstad et al. 1986; Segalstad & Ohmoto 1986; Segalstad
1996; Segalstad 2000).
Thermochemical modelling shows that the hydrothermal solution was poor in sulphur ( G S # 0.01 m) and rich in carbon ( G C $ 0.1 m). Chemical reactions between a saline solution of this kind and pyrrhotite in the fahlbands at the temperature of minimum calcite solubility (near 250 E C) consumed protons, raising the pH of the localized, now closed-system fluids in the tension gashes, leading to the deposition of first argentite and then native silver with “coalblende”. When pyrrhotite stability had been reached in the mineralizing sequence, equilibrium with wall rocks was reached, and the mineralization of silver minerals ended (Segalstad 2001).
The Kongsberg hydrothermal system could have acquired its silver through leaching of the black shales of the
Oslo Region, like it acquired its carbon, either locally from the shales overlying and falling into the opening

60
N G F A bstracts and Proceedings, N o. 2, 2003 fractures (Frøyland & Segalstad 1992), and/or by the hydrothermal fluids flowing along the sub-Cambrian conglomerate in the Oslo Region (Segalstad & Ohmoto, 1990). In a 50 m thick prism of such black shale, overlying the 5 x 20 km Kongsberg hydrothermal system area, it will just be necessary to leach one tenth of the silver available in the shale in order to deposit the 2,000 tons of silver exploited from the mines. The solutions would require only some 3 ppm of the available FeS in the fahlbands to deposit the 2,000 tons of known silver once present in the mines (Segalstad 1996).
The Fen circular carbonatite complex (Fig. 2) is composed of carbonatites and alkaline silicate rocks, associated with breccias, explosion pipes, and kimberlitic-lamprophyric dikes (Ramberg & Barth 1966). The Fen Complex, which has a diameter of about 2 km, was among the first carbonatite – peralkaline silicate rock complexes to be recognized and studied.
Iron mining (to 200 m depth) had been undertaken in the Fen area from 1652 to 1927, yielding nearly one million tonnes of iron ore: hematite in veins and lenses in “rødberg” (iron-bearing carbonate rocks). This ore deposit has been described by Vogt (1910, 1918) and Andersen (1983). The ore contained 50% Fe, 0.45% P, 1-2% Mn, and
0.3% S. Vogt found that the iron-bearing carbonate rocks were separated from the surrounding Precambrian rocks by a fault. He then assumed that a block of Cambro-Silurian limestone had subsided into the basement, and later had been subjected to thorough metasomatism.
In May 1918 Dr. V.M. Goldschmidt discovered the nepheline-bearing alkaline igneous rocks in the Fen area.
He reported this discovery to Professor W.C. Brøgger, and they then initially assumed that the Fen area represented an “outlier” of the Oslo Region. They decided that they needed to do a thorough investigation of the area, and they went together to Fen to make a geological map. This map is published in Brøgger’s (1921) comprehensive petrographic and field description of the Fen area. During this work, Brøgger realized that the
Fen area was not an “outlier” of the Oslo Region, but a subsurface volcanic explosion centre, a diatreme, of possible Eocambrian (= late Precambrian) age. He also realized that the carbonate rocks of the Fen area were true to the carbonatite described by Högbom (1895) from the similar Alnö Complex near Sundsvall, Sweden.
The rocks in the Fen area are rich in carbonates and phosphates, which are good fertilizers for plants. Therefore the vegetation makes it difficult to make a good geological map of the area. It is being said that Brøgger and
Goldschmidt investigated the peculiar Fen rocks in stone fences, assuming that their stones were acquired locally, and named the rocks after the nearest farm or geographical feature: vibetoite, melteigite, kamperite, juvite, rauhaugite, søvite, damkjernite, fenite, and more. N.L. Bowen described W.C. Brøgger in his obituary as a glowing patriot, with an ambition that as many different rock-types as possible should be named from Norwegian localities, adding that it would be good if all patriotism took an equally harmless form (Hestmark 1999).
The petrography of the Fen area is summarized by Sæther (1957), Fig. 2:
Fenite (named from the Fen farm) is a syenitic rock surrounding the Fen area, formed from gneiss by alkali
(mainly sodium) metasomatism, a process called fenitization .
Vibetoite (= vipetoite) – melteigite – ijolite – urtite is a series of mafic silicate rocks consisting of pyroxene, amphibole, and nepheline.
Damkjernite (= damtjernite) is a hypabyssal utrabasic alkaline rock carrying large phenocrysts of magnesian biotite (phlogopite) in a fine-grained groundmass of titaniferous clinopyroxene, amphibole, mica, nepheline, microcline, albite, calcite, magnetite, plus lzerzolite nodules (Griffin 1973).
Søvite is a carbonatitic rock with mainly calcite and minor silicates.
Rauhaugite is a carbonatitic rock with mainly ankerite.
Rødberg (= “red rock”) is a carbonatite rock with calcite and ankerite in varying proportions, with red colour from finely dispersed hematite. When hematite at many places becomes the main constituent, the rock has been mined as an iron ore.

N G F A bstracts and Proceedings, N o. 2, 2003
61
Fig. 2. Geological map of the Fen Complex, Telemark; modified from Sæther (1957) and Ramberg & Barth (1966).
Brøgger (1921) proposed that the silicate magma had assimilated a Precambrian carbonate rock at great depth in the Earth’s crust, so that the then formed carbonatite magma could exsolve at lower pressure and erupt together with the silicate magma. For the rauhaugite and the rødberg he assumed a metasomatic formation from søvite. Bowen (1924, 1926) rejected totally the idea of a carbonatitic magma. Instead he maintained that the søvite had been formed through metasomatic replacement of silicate by calcite.
In the important textbook “The Evolution of the Igneous Rocks” Bowen (1928) states: There is nothing in the manner of their [primary igneous carbonate rocks] occurrence to warrant belief in their formation by simple consolidation of a magma, nor is there in “carbonate dikes” . Although Bowen (1928) recognized liquid immiscibility in magmas, he wrote elsewhere in his book: It is apparently not realized by many petrologists that unmixing is not a mysterious process whereby one liquid separates instantaneously into two masses of liquid of different composition and with but one common surface boundary surface. Unmixing is, on the contrary, a manifestation of phase equilibrium just as is crystallization.
Bowen’s criticism of Brøgger’s work on the Fen area began a controversy that persisted for nearly 40 years.
Brøgger’s views of a carbonatite magma was defended by von Eckermann (1942, 1948) in his publications from
20 years of studies on the Alnö Complex. Sæther (1957) favoured, however, a “peri-magmatic hydrothermal” model for the formation of the søvite and its accompanying carbonate rocks, where the carbonate solutions had exsolved from deeper parts of the magma, whose parent magma had a kimberlitic composition. Bergstøl &
Svinndal (1960) preferred a metasomatic model for the carbonatite petrogenesis.
Wyllie & Tuttle (1960) provided experimental evidence for the existence of magmatic carbonatite, even at low temperatures and pressures. Despite this, a magmatic origin for carbonatites was not universally accepted until carbonate lavas were witnessed erupting from the O ldoinyo Lengai volcano in Tanzania (Dawson 1962). In
Tuttle & Gittins’ book “Carbonatites”, Barth & Ramberg (1966) could accentuate the magmatic carbonatitic nature of the søvite without protests. Mitchell & Brunfelt (1975) argued that all the Fen rocks are placed in a

62
N G F A bstracts and Proceedings, N o. 2, 2003 petrogenetic scheme in which a carbonated nephelinite magma underwent liquid immiscibility, differentiation
(fractional crystallization), and volatile transport.
Gravity studies by Ramberg (1973) showed a large positive gravity anomaly over the Fen area, indicating that dense rock masses occur in a pipe-like fashion to at least 15 km depth, possibly more. The deep origin is also supported by the occurrence of lherzolite nodules (xenoliths) in damkjernite. These nodules crystallized at or near the base of the continental crust at 33-34 km depth; pressure 10-13 kbar; temperature 1,200-1,250 E C
(Griffin 1973).
so they concluded that søvite represented the magma from which the other rocks of the complex were derived.
Griffin & Taylor (1975) estimated that the groundmass of the damkjernite magma formed from partial melting
1,200-1,250 E C. They pointed out that damkjernite has no chemical resemblance to kimberlite, but is similar to alnöite, monchiquite, and ouachitite.
Tom Andersen presented in the 1980s a number of papers on the Fen area based on modern petrologic methods, compiled in his doctor’s dissertation (Andersen 1987; containing 12 papers). His microstructural analyses of carbonatites and iron ore suggested that the rødberg and its associated hematite ores formed by postmagmatic oxidation of pre-existing ankeritic ferrocarbonatite and magnetite – pyrite ores. The magnetite ores were themselves formed by hydrothermal processes during an early stage of postmagmatic alteration of the ferrocarbonatite intrusion. Thermodynamic modelling of the metasomatic processes suggested by Brøgger (1921) and Sæther (1957), showed it to be impossible to introduce the required amount of ferric iron during oxidation, due to low ferric ion solubility (Andersen 1983). O and Sr isotopes further showed that magmatic fluids were not involved in the metasomatic alteration of the ferrocarbonatite. Trace-element and isotope data indicated that circulation of groundwater may have played a substantial role in the post-magmatic alteration of the carbonatites.
The metasomatism caused a major volume reduction of the carbonatite, making some elements mobile and others immobile (Andersen 1984).
The discovery of cognate xenoliths of apatite cumulate in carbonatites in the Fen Complex (Andersen 1986), allowed the fluid evolution of the carbonatite magma to be studied. It was shown that the apatites were formed from a magnesian calcite carbonatite melt in the middle crust at pressure $ 4 kbar and temperature $ 600 E C.
Andersen (1987) further presented Sr, Nd, C, O, and Pb isotopic data from the Fen rocks, pointing out a depleted mantle source and contamination of the magmas from the local continental crust.
Andersen (1987) found the petrogenetic model by Mitchell & Brunfelt (1975) too simple to fit the observed major-, trace-, and isotopic-compositional features observed in the carbonatites. Andersen’s (1987) new petrogenetic model for the Fen Complex illustrates the importance of processes such as fractional crystallization, liquid immiscibility, volatile transport, and crustal contamination for the evolution of the carbonatite magma; the carbonatite-forming trends represent different lines of descent originating from a common parent magma, probably of nephelinitic composition:
1) Immiscible peralkaline calcite carbonatite and nepheline syenite liquids evolved after shallow fractionation of ijolitic magma.
2) Alkaline biotite-amphibole søvite and dolomite carbonatite formed from a Mg-bearing calcitic protomelt by fractionation and wall-rock interaction in the middle to shallow crust. The primitive carbonatite melt was immiscible with a mafic silicate melt at pressure $ 4 kbar.
3) Heterogenous silicate-rich ankeritic ferrocarbonatite formed from a melt which was immiscible with the mafic silicate melt in the shallow crust.
4) Dikes and veins of silicate-poor ankeritic ferrocarbonatite formed from “hydrosaline melt”, which was a byproduct of process 2).

N G F A bstracts and Proceedings, N o. 2, 2003
63
Age determinations of the Fen area rocks have given ages ranging from 523 to 601 M a (M eert et al. 1998). The most recent age determinations of the Fen Complex were given as 539 ± 14 Ma (whole-rock Pb-Pb isochron) and 550 ± 7 Ma (Rb-Sr mineral isochron) by Andersen (1987). Dahlgren (1994) investigated more than 50 macrocrysts from 10 of these satellites yielded a Rb-Sr isochron age of 578 ± 24 Ma. Meert et al. (1998) found
Dahlgren (1994) found that a dike similar to the Fen damkjernites at Bolkesjø, 51 km N of Fen, was emplaced at 324 ± 4 Ma, dating the initiation of the Oslo Rift. Anthony et al. (1989) calculated Nd model ages for the
Skien nephelinites in the Oslo Rift, obtaining 600 to 800 Ma for magma separation from the depleted mantle source, only slightly older than the age of the Fen Complex. From similarities in REEs and the isotopes of Sr,
Nd, and Pb, Anthony et al. (1989) proposed that the carbonate fluids invading the mantle and generating the Fen magmatism may have left a localized LREE- and Sr-enriched, Rb-depleted residue, which became the source of the Skien nephelinitic magma.
The discovery of niobium deposits associated with the carbonatites led to renewed interest in the economic geology of the Fen area from the late 1930s, leading to the opening of the Søve niobium mines (mined 1951 to
1978). The niobium is contained in the pyrochlore mineral koppite. Columbite occurs as pseudomorphs after koppite.
Open-pit mining started on the Cappelen ore, until the ore body terminated against a fault. The continuation of the ore was found about 60 m below, and a shaft was sunk for mining purposes. The Cappelen deposit contains
0.5% pyrochlore, 7% apatite, 3% magnetite, 1% pyrite, and 80% calcite. The silicate minerals are alkali amphibole, tremolite, mica, and zoisite. Accessory minerals are fluorite, topaz, zircon, and barite. The Hydro deposit had too low niobium content to be mined. A tunnel was therefore drifted 900 m southwards into the central part of the large søvite body (Tufte mine). This ore had a higher dolomite content than the Cappelen ore.
years of operation the production was 30,000 tonnes of ore annually from the Tufte mine, and 100,000 tonnes annually from the Cappelen quarry. Falling prices and increasing production costs forced the closure of the niobium mining at Fen (Bugge 1978).
The niobium content decreases in the carbonatites from søvite through rauhaugite to rødberg, whereas the contents of yttrium, REE, and thorium increase. Dahlgren (1983) carried out gamma-ray mapping of the Fen area and showed a significantly increased natural radioactivity, primarily from thorium, in the same sequence; the rødberg showed a gamma-ray radiation intensity of more than 400 µR/h. Inside the old iron mining works up to 1,000 µR/h have been measured. Søvite contains 1 to 50 ppm Th; rauhaugite up to 2,000 ppm Th; rødberg up to 3,000 ppm Th.
The rødberg contains 2.5 to 4 wt.% REE minerals (dominating: monazite, synchysite, and parisite). There are large rock reserves containing on the average 200 ppm Y; 2,000 ppm La; 300 ppm Sm; 1,400 ppm Th. Thorium is further enriched in the hematite ore, containing about 0.2 wt.% Th (Bugge 1978) and about 1 wt.% LREE
(Bergstøl & Svinndal 1960).
Mitchell & Brunfelt (1974) reported Sc contents in the Fen area rocks. Averages were: urtite 1.8 ppm; ijolite 6.6
ppm; vibetoite 5.2 ppm; silicocarbonatites 3.2 ppm; søvite 3.5 ppm; damkjernite 24 ppm; rauhaugite 15 ppm; rødberg 61 ppm.
Åmli (1977) carried out an electron microprobe study of the Fen area rauhaugite and rødberg carbonatites. Also he did whole-rock Sc analysis of 128 samples from 700 m of diamond drill cores. He found small (2 to 3 µm)

64
N G F A bstracts and Proceedings, N o. 2, 2003 grains of thortveitite and up to 160 µm long anhedral to euhedral grains of scandium-bearing columbite (0.38
2 3
32-131 ppm Sc, with averages 52 ppm Sc for rauhaugite, and 73 ppm Sc for rødberg. He concluded that carbonatites may constitute a potential source for scandium exploitation.
We have recently analyzed two samples of rødberg from the Fen area by instrumental neutron activation for Sc, Th, U, and eight rare-earth elements. The samples are from the collections of the Geological
Museum, University of Oslo. The results below are in ppm of the elements:
Cat. No. 57042 (Gruvedalen; collected 1920 by Brøgger):
Sc 167; Th 550; U 8.4; La 1,110; Ce 2,080; Nd 879; Sm 188; Eu 75.2; Tb 22; Yb 14.6; Lu 2.21
Cat. No. 57043 (Stinta, SW of Rauhaug; collected 1919 by Brøgger & Goldschmidt):
Sc 10.8; Th 136; U 10.4; La 260; Ce 424; Nd 200; Sm 52.2; Eu 15; Tb 5.6; Yb 7.7; Lu 1.19
A high Sc content is correlated with high contents of the light REEs and Th. Polycrystals of parisite / synchysite are reported from these rocks (Neumann 1985).
by lepidolite and cassiterite, Fig. 3 (Segalstad & Eggleston 1993a). Hydrothermal veins carrying molybdenite and cassiterite also occur in the area (Bugge 1963, Bergstøl et al. 1985). The mineralogy of the pegmatites was first described by O ftedal (1942). The occurrences were mined for mica, amazonite feldspar, and molybdenite during the 2nd World War.
Fig. 3. Geological map of the Tørdal pegmatite area, Telemark, m odified from Segalstad & Eggleston (1993a). Stippled lines separate zones with pegmatites dominated by green amazonite, white K-feldspar, and pink K-feldspar. Notable mineral occurrences are named and indicated by circles. The grid is based on the UTM system, European Datum 1950.

N G F A bstracts and Proceedings, N o. 2, 2003
65
Several hundred of these pegmatite bodies, of age approximately 900 Ma (Neumann 1960a), occur throughout a large area of a Precambrian metamorphosed supracrustal sequence (Mitchell 1967). Many of the pegmatites contain amazonite as a major mineral, as well as numerous rare minerals (see the list below). Cleavelandite and quartz have replaced earlier minerals. The amazonite feldspar is found to have been replaced by white and pink alkali feldspar.
Segalstad & Eggleston (1993a; and Eggleston & Segalstad in prep.) found that the colour of the K-feldspar within the Tørdal pegmatite district exhibits a distinct regional zoning, which can be divided into three main, east-west trending zones: an amazonite-dominated zone, a white K-feldspar-dominated zone, and a pink Kfeldspar-dominated zone (Fig. 3). The amazonite-dominated zone is the farthest from the contact to the Tørdal granite, and consists of numerous, small pegmatites (less than 3 m wide and rarely exceeding 100 m in length) in which amazonite is the dominant K-feldspar. Important exceptions are the Høydalen and Skarsfjellet pegmatites (100 to 200 m long and 20 to 30 m wide). Pegmatites in this zone contain most of the rare minerals that occur in the district: beryl and cassiterite are most common in pegmatites within this zone; blue bazzite (the scandium analogue of beryl) and blue beryl occur in the western part; beryl from Skarsfjellet and eastwards to
Høydalen is typically light yellow, light green, red, and/or colourless.
The boundary between the amazonite-dominated zone and the white K-feldspar-dominated zone is gradational over large distances. Within that boundary zone, amazonite is progressively replaced by white K-feldspar and by cleavelandite. The cleavelandite forms veinlets where the amazonite is first bleached white and then replaced by cleavelandite. In the white K-feldspar-dominated zone, amazonite locally occurs as relict cores of large Kfeldspar crystals. Pegmatites here contain few rare minerals; beryl and gadolinite may occur sporadically.
The transition between the white- and the pink K-feldspar-dominated zone of pegmatites is gradational over a large distance. Presence of pink K-feldspar, not the absence of white K-feldspar, is diagnostic of this zone. These pegmatites tend to be very large, and consist predominantly of very coarse granitic zone with some graphic granite and only small zones of coarse core-wall material. Core quartz is abundant, and typically forms small pods and veinlets scattered throughout the pegmatite. Molybdenite is more common here than in either of the other two zones; the largest known molybdenite occurrences are near Stormyr and Kleppe.
Both simple and complex pegmatites, crosscutting or foliation-aligned, are found in the Tørdal area. Using the pegmatite terminology defined by Uebel (1977), the complex pegmatite type has a thin granitic border zone , followed by a graphic granite zone . The following core-wall zone contains most of the rare minerals. The innermost core zone quartz may extend as veins through the pegmatite bodies and further into the wall-rocks.
Pegmatite borders to the amphibolite wall-rocks are sharp, and show plastically deformed wall-rocks around the pegmatite bodies. Mica in the core-wall zone may be strongly bent and curved around the pegmatite core. These findings indicate that large pressures must have developed in the pegmatite core.
Primary fluid inclusions are extremely hard to find in the pegmatite minerals of the zones following the graphic granite and early core wall zones. Most of the minerals contain numerous internal cracks, as if most of the primary inclusions had high internal pressures and later decrepitated forming cracks. Three types of fluid inclusions occur in these pegmatites: Type 1 comprises primary inclusions containing aqueous fluid and CO
2
(L+V) found in quartz of the graphic-textured zone. Type 2 comprises multi-phase, liquid-vapour-solid inclusions that occur in topaz associated with the cleavelandite replacement stage. Liquid-vapour homogenization occurs at 213 to 270 E C, and the unidentified daughter minerals dissolve at 460, 510, and 580 E C; 4 eq. wt. %
NaCl. Type 3 inclusions are exclusively secondary liquid-vapour inclusions with halite and sylvite daughter minerals; T = 70 to 400 E C; 2 to ~ 40 eq. wt. % NaCl. The fluid-inclusion microthermometry on topaz, together with available mineral-stability data, indicate that the Tørdal granite and its early pegmatites were formed at 8 to 10 km depth at a pressure of 2.4 ± 0.4 kbar and a temperature of approx. 600 E C. Later parts of the pegmatites formed after water exsolved from the pegmatitic melt, exerting an increased pressure, up to a total of 4-5 kbar.
The pegmatite evolved in such a way that the pegmatitic melt cooled slowly along the solidus pressure /

66
N G F A bstracts and Proceedings, N o. 2, 2003 temperature curve for a fluorine-containing, water-saturated granitic melt. "Ordinary" minerals crystallized first from the aqueous silicic melt, and the remaining hot silicic aqueous solution was enriched in rare elements from which rare minerals formed. The high-pressure fluid inclusions would not be easily preserved at the Earth’s surface, where the inclusions would decrepitate; this explains the absence of beryl (and other minerals) in gemstone quality from Tørdal (Segalstad & Eggleston 1990, 1993a, 1993b).
The pegmatitic granite dikes are most likely related to the large non-foliated granite body informally referred to as the Tørdal granite. This granite is fine to medium grained and consists of orthoclase (50-60%), plagioclase
(15-20%), interstitial quartz (10-15%), and late, interstitial biotite (1-2%). Zircon, allanite, and opaque minerals are accessory phases. Numerous almandite-bearing amphibolite xenoliths occur in the granite, showing different degrees of assimilation. At places the granite is almandite-bearing, the garnets being inherited from the assimilated amphibolites. In the Tørdal area the Nissedal supracrustal sequence consists predominantly of amphibolites, quartzofeldspathic schists, granitic gneisses, and quartzite. Mitchell (1967) considered the amphibolites in the Nisservann area, west of the Tørdal area, to be mafic and ultramafic metavolcanic rocks, including meta-pyroclastic volcanic sedimentary rocks, of basaltic and andesitic composition. In the Tørdal area, sedimentary rocks are abundant in the supracrustal sequence. In a number of aspects the Tørdal granite resembles an S-type granite, formed through anatexis of sedimentary rocks (White & Chappell 1977), known to be associated with tin deposits of S-type magmatic heritage (Burnham & Ohmoto 1980).
Bergstøl et al. (1985) showed that micas in the supracrustal rocks, the granite, and the pegmatites all show elevated values for tin (muscovite 350 to 2,000 ppm Sn; biotite 25 to 835 ppm Sn) in this Norway’s only tin province. They also found elevated values in the rocks of the area for Sc, Rb, Li, W, F, Be, and Mo.
Oftedal (1943) found the following Sc contents in Tørdal minerals:
Høydalen: muscovite up to 700 ppm; beryl up to 300 ppm; cassiterite 100 ppm Sc.
Skarsfjell: muscovite up to 400 ppm; zinnwaldite 2,000 ppm; topaz 3 ppm.
Biotite from amphibolite wall-rock 100 ppm; from hornblende gabbro at Høydalen: amphibole 50 ppm, and biotite 5 ppm; biotite from Tørdal granite 30 ppm Sc.
Bergstøl & Juve (1988) found the following Sc and Sn contents (averages in ppm) in the Tørdal rocks:
Tørdal granite: Sc 4.5; Sn 0.4
Granite pegmatite within the Tørdal granite: Sc 1.8; Sn < 1
Amphibolite inclusions in the Tørdal granite: Sc 17.7; Sn 4.0
Amphibolites in the Tørdal area: Sc 31.4; Sn 3.4
Hornblende-biotite gneiss: Sc 20.0; Sn 6.8
Leptite (rhyolitic tuff): Sc 7.8; Sn 15.5
The pegmatites of the Tørdal area show local enrichment in Sn and/or Sc. Lower contents of Sn and Sc in the
Tørdal granite, and relatively high contents of these elements in the supracrustal rocks, made Bergstøl & Juve
(1988) to conclude that some of the Sn and a major part of the Sc found in amazonite-cleavelandite pegmatitetes may have had their source in the supracrustal rocks in the area. They suggested that the “tin in the pegmatite veins may partly be due to contamination from the supracrustal rocks during the emplacement of the crosscutting veins”, similarly proposed for Sc by Goldschmidt (1934) for the thortveitite-bearing pegmatites of Evje-
Iveland.
We agree to this model, but want to add that the granite contains a large number of inclusions of supracrustal rocks, mainly amphibolite. These xenoliths may be looked upon as “restites” in a granite formed by anatexis, in the sense of White & Chappel (1977). Complete or partial melting of the Sn- and Sc- enriched supracrustal rocks would lead to an enrichment of these elements (plus other incompatible elements, with large ionic radii and high ionic charges) in the volatile-rich last pegmatitic melt of the granitic magma. This last aliquot of aqueous melt was rich in fluorine and chlorine, which would act as strong complexers for many of these elements, including Sc. The only mafic mineral in the granite which could adopt some Sn and Sc from this anatectic process would be biotite, which indeed shows elevated values for these two elements, supporting our

N G F A bstracts and Proceedings, N o. 2, 2003
67 supposition. See more on the formation of scandium minerals below under the part “Scandium geochemistry; the formation of thortveitite” (in the Evje-Iveland Pegmatite Area chapter).
The occurrence of Sc minerals in the Heftetjern granite pegmatite was first described by Bergstøl & Juve (1988) would make this a new mineral species (Raade & Kristiansen 2002a, 2002b). Further investigations have identified the two pyroxenoid minerals cascandite and scandiobabingtonite, in part as a fibrous intergrowth
(Raade & Erambert 1999). Heftetjern is the type locality for the new mineral species kristiansenite, a triclinic
2 7 2 6 in progress on another new, Sc-dominated mineral species from Heftetjern, a milarite-related mineral with end-
12 30 thortveitite, bazzite, Sc-dominated milarite, and kristiansenite is described by Raade & Bernhard (abstract in this
With so much Cs in the structural channels, it is likely that Li substitutes for Be in tetrahedral framework positions. This possibility will be explored at a later date. It should be noted that so far, no Li-bearing minerals except probably zinnwaldite have been identified from the Heftetjern pegmatite. The mineralogy and geochemistry of the Heftetjern pegmatite have been briefly treated by Raade & Kristiansen (2002a, 2002b). A summary of the Heftetjern mineralogy was also given by Eldjarn (2002). The minerals so far identified from the
Heftetjern pegmatite are listed below, with reference to the papers where the minerals are first mentioned.
Černy (1992) has established three families of rare-element pegmatites: (1) the LCT family which typically carries Li, Rb, Cs, Be, Sn, Ga, Ta > Nb, (B, P, F); (2) the NYF family which is marked by a Nb > Ta, Ti, Y, Sc,
REE, Zr, U, Th, F signature; (3) a mixed LCT + NYF family. The Heftetjern pegmatite belongs to the latter, mixed category. It is especially characterized by Cs, Be, Sc, Sn, and Ta. Since the micas have not been studied in detail, we do not know about their contents of Li and F.
Minerals from the Heftetjern granite pegmatite, Tørdal, Telemark, Norway:
Actinolite (Raade & Kristiansen 2000a)
Albite (in part: variety cleavelandite) (Juve &
Bergstøl 1988)
Allanite-(Ce) (Juve & Bergstøl 1988)
Anatase or brookite (Raade & Kristiansen 2000a)
Apatite (Raade & Kristiansen 2000a)
Augite (Raade & Kristiansen 2000a)
Bavenite (Bergstøl & Juve 1990)
Bazzite (Bergstøl & Juve 1988, Juve & Bergstøl
1990)
Bertrandite (Juve & Bergstøl 1988)
Beryl (Juve & Bergstøl 1988)
Biotite (?) (Raade & Kristiansen 2000a)
Bismuth (Kristiansen, pers. comm.)
Bismutite (Kristiansen, pers. comm.)
Calcite (Raade & Kristiansen 2000a)
Cascandite (Raade & Erambert 1999)
Cassiterite (Juve & Bergstøl 1988)
Cerussite (Raade & Kristiansen 2000a)
Clinochlore (Raade & Kristiansen)
Epidote, cerian (Raade & Kristiansen 2000a)
Ferrotantalite (Raade & Kristiansen 2000a)
Fluorite (Juve & Bergstøl 1988)
Gadolinite (Juve & Bergstøl 1988)
Galena (Raade & Kristiansen 2000a)
Goethite (Kristiansen, pers. comm.)
Hellandite-(Y) (Raade & Kristiansen 2000a)
Helvite (Raade & Kristiansen 2000a)
Hingganite-(Y) (Raade & Kristiansen 2000a)
Ilmenite, manganoan (Kristiansen, pers. comm.)
Ilmenorutile (Raade & Kristiansen 2000a)
Ixiolite, scandian (Bergstøl & Juve 1988)
Kainosite-(Y) (Raade & Kristiansen 2000a)
Kristiansenite (Raade et al. 2002)
Magnetite (Bergstøl & Juve 1990)
Microcline (in part: variety amazonite) (Juve &
Bergstøl 1988)
Microlite (Juve & Bergstøl 1988)
Milarite (Raade & Kristiansen 2000a)
Milarite, scandium analogue (Hawthorne 2002)
Milarite, yttrium analogue (Kristiansen, pers. comm.)
Molybdenite (Kristiansen, pers. comm.)
Monazite-(Ce) (Bergstøl & Juve 1990)
Muscovite (Raade & Kristiansen 2000a)
Phenakite (Eldjarn 2002)

68
N G F A bstracts and Proceedings, N o. 2, 2003
Plumbomicrolite (Raade & Kristiansen 2000a)
Pyrite (Raade & Kristiansen 2000a)
Pyrochlore, scandian (Bergstøl & Juve 1988)
Opal (Raade & Kristiansen 2000a)
Quartz (Juve & Bergstøl 1988)
Rynersonite (?) (Eldjarn 2002)
Scandiobabingtonite (Raade & Erambert 1999)
Schörl (Raade & Kristiansen 2000a)
Spessartine (Bergstøl and Juve 1988)
Stilpnomelane (Raade & Kristiansen 2000a)
Thortveitite (Eldjarn 2002, Raade & Bernhard, this volume)
Titanite (Raade & Kristiansen 2000a)
Yttrobetafite-(Y) (Juve & Bergstøl 1988)
Yttropyrochlore-(Y) (Juve & Bergstøl 1988)
Yttrotantalite-(Y) (Raade & Kristiansen 2000a)
Zinnwaldite (Bergstøl & Juve 1990)
Zircon (Juve & Bergstøl 1988)
Paragenetic sequence for the Heftetjern pegmatite (after Eldjarn 2002); scandium minerals in bold:
MINERAL PRE-SCANDIUM STAGE SCANDIUM-RICH STAGE POST-SCANDIUM STAGE
Quartz ——————————— ——————
Amazonite ———————————
Albite ————————————————————————
Beryl ————————
Phenakite ——
Bazzite ——————————
Scandiobabingtonite / cascandite ————————————
Thortveitite —————
Kristiansenite ————————————
Ixiolite ———————————
Microlite / yttropyrochlore —————————— ———————
Ferrotantalite ———————
Helvite ——————————
Milarite ———————————————
Bertrandite ————————————
Spessartine ————————————
Allanite-(Ce) ——————————————
Titanite ————————————
Rynersonite ————————
Stilpnomelane ——————————————
Clinochlore ————————————
The complex granite pegmatites at Høydalen occur in an amphibolite of assumed meta-gabbro origin (Oftedal
1942). In the upper quarry the pegmatite dike is about 5 m thick, with E-W strike and varying dip close to vertical. Borders to the wall-rocks are sharp, showing ductile deformation of the wall-rocks from volume expansion of the pegmatite. Next to the wall-rock may be a 1-2 cm thin, fine-grained granitic zone, followed by a coarser graphic-granite zone, the core-wall zone, and the core zone. The cleavelandite pegmatite occurs mainly in the core-wall zone.
The complex granite pegmatite dike in the lower quarry is about 4 m thick, with SE-NW strike, with a steep and varying dip towards NE. This complex pegmatite is not as symmetrically zoned as the dike in the upper quarry, but otherwise almost identical. The cleavelandite pegmatite is less developed here (Oftedal 1942).
The most recent listing of the minerals from the Høydalen pegmatites was published by Kristiansen (1998).
Below is given a complete list of the Høydalen minerals with reference to the papers where the minerals are first

N G F A bstracts and Proceedings, N o. 2, 2003
69 mentioned. Høydalen is the type locality of tveitite (Bergstøl et al. 1977), which has a trigonal, fluorite-derived named for John Peder Tveit (1909-1978), father of the present owner of the Høydalen and Heftetjern deposits,
Kaj-Peder Tveit.
The Høydalen pegmatites belong to the mixed LCT + NYF rare-element pegmatite family of Černý (1992). In contrast to the Heftetjern pegmatite, they are marked by a distinct Li, F, Y, Ta signature. The elements Be and
Sn are common for both occurrences. Černý & Ercit (1985) says about the Høydalen pegmatite area: “Evidently a strong Y, REE signature of the whole pegmatite district persists through advanced fractionation into the Li,
F-enriched, lepidolite-bearing pegmatite type”. Oftedal (1970) found 0.2 wt.% Li in a light red beryl from
Høydalen.
Discrete Sc minerals have not been found in Høydalen. Microscopic inclusions (0.05 to 0.37 mm) of ixiolite or contents of the Høydalen minerals has so far been undertaken. With regard to the proposed transport mechanism of Sc as a fluoro complex (Gramaccioli et al. 2000), it is interesting to note the absence of Sc minerals in the
Høydalen pegmatites, which are rich in fluorine-bearing minerals, as compared to the abundance of Sc minerals in the Heftetjern pegmatite, which apparently is quite poor in fluorine.
Minerals from the Høydalen granite pegmatites, Tørdal, Telemark, Norway:
Albite (in part: variety cleavelandite) (Oftedal
1942)
Allanite-(Ce) (Raade et al. 1993)
Axinite (Raade et al. 1993)
Bastnäsite-(Ce) (Raade et al. 1993)
Bavenite (Raade et al. 1993)
Beryl (Oftedal 1942)
Calcite (Raade et al. 1993)
Cassiterite (Oftedal 1942)
Cerianite-(Ce) (Bergstøl et al. 1977)
Clinozoisite / epidote (Raade et al. 1993)
Fergusonite-(Y) (Raade et al. 1993)
Fluocerite-(Ce) / tysonite (Sverdrup et al. 1965)
Fluorite (Oftedal 1942)
Fluorite (variety yttrofluorite) (Sverdrup et al.
1965, Sverdrup 1968)
Gadolinite-(Y) (Oftedal 1942)
Goethite / lepidocrocite (Kristiansen 1998)
Hingganite-(Y) (Juve & Bergstøl 1997)
Kainosite-(Y) (Bergstøl et al. 1977, Raade et al.
1993)
Kamphaugite-(Y) (Raade & Brastad 1993, Raade et al. 1993)
Kuliokite-(Y) (Raade et al. 1993)
Laumontite (Raade et al. 1993)
Lepidolite (Oftedal 1942, Bailey & Christie 1978,
Rule et al. 1987)
Microcline (in part: variety amazonite) (Oftedal
1942)
Microlite (Oftedal 1942)
Milarite (Kristiansen 1998)
Molybdenite (Segalstad & Eggleston 1993a)
Monazite-(Ce) (Oftedal 1942)
Muscovite (Oftedal 1942); Mn-bearing muscovite-2M (Segalstad & Eggleston 1993a)
Quartz (Oftedal 1942)
Pyrite (Kristiansen 1998)
Spessartine (Oftedal 1942)
Synchisite-(Y) (Kristiansen 1998)
Tantalite (
erný & Ercit 1985)
Tengerite-(Y) (Raade et al. 1993)
Thalenite-(Y) (Kristiansen 1998)
Topaz (Oftedal 1942)
Tourmaline (magnesian schörl?) (Raade et al.
1993)
Tveitite-(Y) (Bergstøl et al. 1977)
Wodginite (or ixiolite?) (Raade et al. 1993)
Xenotime-(Y) (Raade et al. 1993)
Zircon (variety alvite) (Oftedal 1942)
Yttrotantalite-(Y) (Oftedal 1942)
In fluid inclusions:
Halite (Segalstad & Eggleston 1993a)
Sylvite (Segalstad & Eggleston 1993a)

70
N G F A bstracts and Proceedings, N o. 2, 2003
Below is shown a generalized paragenetic sequence for the scandium-mineral-free complex pegmatites of the
Tørdal area, like those at Høydalen. Very rare minerals such as tveitite, yttrofluorite, fluocerite, etc., are included in the notation REE fluorides; Ksp = K-feldspar. Note that the dimensions of the different zones are not reflected from this illustration, e.g., the dimensions of the cleavelandite and the alteration zones are grossly exaggerated in this illustration for typographic reasons (after Segalstad & Eggleston 1993a, 1993b; Eggleston & Segalstad in prep.):
COMPLEX PEGMATITE ZONES
---------------------------------
MINERAL WALL GRAPHIC CORE-WALL CORE CLEAVELANDITE ALTERATION
Quartz ———————————————— — — — — — ——————— — — — — — — —
Amazonite — — — —————————————————————
White Ksp ——————————————————————————— ——————————
Pink Ksp ———————————————
Plagiocl. —————
Cleavelandite —————————————
Biotite ————— — — — — — — — — — — — — —
Muscovite — — — ——————————— — — — — — — — — — — — — — — — ———
Zinnwaldite —————————————
Lepidolite —————————————
Beryl ——————————— — — — — — — — — — —
Topaz ——————————— — — — — — — —
Spessartine — — — — — — — — — — —————————————
Molybdenite ————————————————————————
Cassiterite —————————— — —
Fluorite ———— — — —
Gadolinite — — — — — — —————————————
Yttrotantalite — — — — — — —————————————
REE fluorides — — — — — — —————————————
Monazite —————————————
Tourmaline —————————————
Ferrimolybdite ———
Amazonite is a green variety of microcline with exsolved, white perthite lamellae of albite. Amazonites have cooled extremely slowly (Tibballs & Olsen 1977). The green colour has been ascribed to a high content of Pb, analysed a large number of major, minor, and trace elements from different coloured feldspars from the
Landsverk I pegmatite at Evje (see the 4th day of this field trip), and found no significant difference in these elements to explain the difference in colouring. They concluded that a difference in colours must be caused by physical differences, such as lattice defects or strain, rather than chemical differences. Oftedal (1957) found that amazonite bleached in heating experiments. 100% of the colour was lost at 500 E C, and no colour was lost at
270 E C, leading him to conclude that 270 E C was the maximum temperature at which the green colour could have appeared. However, at Tørdal it appears that the amazonite was formed at a considerable higher temperature, near 600 E C (see above). Oftedal (1954) found as much as 500 (Høydalen) to 1000 ppm (Skarsfjell) Pb in the
Tørdal amazonites. [Such Pb-rich pegmatites being associated with large granite bodies, while Pb-poor (like
Evje-Iveland) are parts of gneiss areas “probably formed by granitisation more or less in situ” (Oftedal 1956).]
Tibballs & Olsen (1977) found from electron microscopy that in Pb-rich feldspars, the Pb was localized to Pbrich grains that did not appear to be Pb feldspar. Rather, subsequent introduction of Na at lower temperatures leads to destruction of pericline twinning, and produces semi-coherent interfaces whose movement generate dislocations, inferred to be the source of the green colouring of amazonite. These dislocations may subsequently

N G F A bstracts and Proceedings, N o. 2, 2003
71 be filled with large-ion grains like the Pb-rich grains, or water molecules. Irradiation experiments on amazonite by Hofmeister & Rossman (1985) showed that the colour depends on the presence of structural water and not on fluid-inclusion water; structural water present was involved in most amazonite colour centres, and suggesting that water plays a catalytic role in amazonite colouration. Amazonite colour can also result from electronic
630 nm) being attributed to the first, and the green colour (600-950 nm; centred at about 720 nm) attributed to be sufficient to produce these colours over geologic time (Hofmeister & Rossman 1985).
Numerous granite pegmatite dikes, rarely exceeding 20 m in thickness and 100 m in length, occur in the Evje-
Iveland amphibolite (Barth 1931, 1947). This amphibolite is 35 km long in the N-S direction, has a maximum
E-W width of 15 km, and is surrounded by gneissic granite and augen-gneisses. Many of the pegmatites have been mined for feldspar to the ceramic industry and for quartz to the metallurgic industry for an extensive period.
Rare minerals have been an important by-product in many of the mines, as well as beryl and muscovite. Today, only the large pegmatite near the farm Li is mined for first-class K-feldspar (“dental spar”); other mines are worked only sporadically.
An ortho-magmatic nickel- and copper-bearing sulphide deposit (pentlandite-bearing pyrrhotite) occurs at Flåt
(2.5 km E of Evje), which produced 3 million tonnes of ore during the period 1870 to 1946. For a long time the
Flåt mine was the largest nickel mine in Europe.
The Evje-Iveland amphibolite consists of three types (Barth 1947): (1) meta-olivine hyperites; (2) meta-norite; and (3) amphibolites of uncertain origin (metamorphic-metasomatic rocks). The meta-gabbroic intrusions were able to plastically deform the gneissic host rocks; the borders between the two rock types are always gradual.
After the mafic intrusions, there was a second magmatic episode with intrusions of monzonite and granite described from the northern part of the area, before the final intrusions of the pegmatites (Pedersen 1975).
The pegmatites occur as sheets or pods, and not like dikes in the magmatic hypabyssal sense. The pegmatites have never been found to make connections to the local felsic magmatic rocks or other pegmatite bodies. Early pegmatite minerals (like beryl, euxenite, and biotite) have often been mechanically deformed (Bjørlykke 1937b).
Taken together with the wall-rock deformation, it appears that a stage of high internal pressure build-up, through unmixing of volatiles from the pegmatite melt, also was a feature of the Evje-Iveland pegmatites, as for the
Tørdal pegmatites (see above).
The complex granitic pegmatites of the Evje-Iveland area usually have an outer granitic zone (fine-grained plagioclase-quartz-muscovite rock); followed by a graphic granite zone (plagioclase + quartz); next a core-wall zone with large sheets of biotite (oriented perpendicular to the borders), in many places with monazite-(Ce), xenotime-(Y), fergusonite-(Y), euxenite-(Y) and other rare minerals; and innermost the core zone quartz with early occurrence of allanite-(Ce), gadolinite-(Y), and beryl (Barth 1947, Bjørlykke 1935, Larsen 2002).
Major minerals of the pegmatites are: quartz and microcline perthite. Common minerals are: plagioclase, biotite, muscovite, garnet (spessartine), and magnetite. Rare minerals are: chrysoberyl, thorite, allanite-(Ce), thalenite-
(Y), gadolinite-(Y), tourmaline, titanite including yttrotitanite, thortveitite, zircon, beryl, bertrandite, fergusonite-
(Y), euxenite-(Y), aeschynite-(Y), polycrase-(Y), samarskite-(Y), columbite, tantalite, yttrotantalite-(Y), monazite-(Ce), xenotime-(Y), ilmenorutile, microlite, betafite, topaz, ilmenite, hematite, molybdenite, pyrite, and chalcopyrite. This area is the type locality of thortveitite, tombarthite-(Y), davidite-(Ce), and the dubious mineral scheteligite.
A second stage in the pegmatite formation involved the introduction of hydrothermal solutions replacing primary
“magmatic” minerals in cavities and along fractures cutting the pegmatites, typically forming white, platy albite
(cleavelandite), amazonite, hematite, fluorite, microlite, tantalite, and hafnian zircon (Frigstad 1999). This latestage pegmatite phase is characterized by enrichment in Ta over Nb, and in Hf vs. Zr.

72
N G F A bstracts and Proceedings, N o. 2, 2003
Fig. 4. Sketch map of the Evje-Iveland pegmatite area. Crosses: pegmatite mines and quarries. Circled crosses: pegmatite field trip stops. Dots: nickel mines. The approximate extent of the Evje-Iveland amphibolite is outlined; note that the border to the gneissic host rocks is always gradual. Modified after Barth (1947). According to Neumann (1961), thortveitite and
Sc enrichment have been found in at least 17 pegmatite mines and quarries scattered over the area.

N G F A bstracts and Proceedings, N o. 2, 2003
73
A review of isotopic age determinations in the Evje-Iveland area by different methods and different authors have been given by Pedersen (1973): age determinations from the pegmatites have given ages from 1,679 to 755 Ma, with most ages clustering around 880 Ma . Not included in the review is a 207 Pb/ 206 Pb age of a gadolinite giving
901 ± 20 Ma (Boudin & Deutsch 1970).
A recent study (Scherer et al. 2001) applied the 207 Pb/ 206 Pb dating method to five fragments of gadolinite from a pegmatite from Evje. The 2 0 7 Pb/ 2 0 6 Pb ages ranged from 890 to 914 Ma, defining a U-Pb discordia upper intercept age of 909 ± 14 Ma. Applying 176 Lu - 176 Hf isotope analysis to three of the gadolinite fragments revealed the most radiogenic hafnium yet measured in a natural sample ( 176 Hf/ 177 Hf = 260 to 272); 99.9% of the 176 Hf was produced by 176 Lu decay in the gadolinite crystal.
A Rb-Sr isotope isochron, based on samples of monzonite and granite from Høvringsvatn in the northern part
1.39 x 10 -11 a (Pedersen 1973). K/Ar isotope age determination of a biotite from the granite gave an age of 845
± 60 Ma (cited as 856 ± 62 Ma in Pedersen 1980), which is closer to the pegmatite ages (Pedersen 1973).
Using a decay constant of 1.42 x 10 -11 a , Pedersen (1980) calculated an age of 1,016 ± 84 Ma for the
Høvringsvatn monzonite and granite. His new analyses give Rb-Sr isotope ages between 945 ± 53 Ma and 853
± 28 Ma for these monzonite and granite intrusions in the northern part of the area. Rb-Sr isotopes of mineral separates indicate, however, that a complete homogenization between the different minerals has not occurred, and that these rock systems have been disturbed by later events, possibly by the regional metamorphism. These
(varying up to 0.7062), indicating an origin from a possible Rb-poor source region (Pedersen 1980).
pegmatite K-feldspar (Stockmarr 1994). From the Rb-Sr isotope age determinations, there is an apparent age spread of 163 million years between the oldest and the youngest age of the granite (Pedersen 1980). Even if the youngest granite age matches the pegmatite K-feldspar age, we do not find that the Rb-Sr isotope ages make a strong enough evidence for a direct genetic link between the igneous rocks and the granite pegmatites occurring in the area. See further discussion of this below.
Stable-isotope data of O, H, and C have been reported by Taylor & Friedrichsen (1983a, 1983b). Quartz from all the primary zones of the pegmatites (from the graphic granite to the core zone) showed normal oxygen isotope different pegmatite minerals reflects the expected values from equilibrium isotope-fractionation factors for pegmatites. δD values (vs. the V-SMOW standard) for pegmatite muscovite range from -37 to -59‰ (mean elsewhere (Taylor & Friedrichsen 1983b).
pegmatite minerals (Taylor & Friedrichsen 1983b). This would not be expected if the pegmatites were magmatic differentiates of these magmatic rocks in the presence of a separate magmatic fluid phase, where the isotopes would distribute themselves among the phases present (Shettel 1978). This has not been considered by Taylor
& Friedrichsen (1983a, 1993b) when claiming that the stable isotope data support an ortho-magmatic origin of all their investigated pegmatites. Partial melting of a granitic gneiss in the presence of a fluid phase already equilibrated with this rock, would not be expected to change its oxygen isotope values much vs. pegmatites crystallizing from the partial melt, assuming that the temperatures of partial melting and pegmatite crystallization are not too far off from each other (and assuming adequate time to obtain isotopic equilibrium between

74
N G F A bstracts and Proceedings, N o. 2, 2003 respectively, for its quartz, K-feldspar, biotite, and magnetite as in the Evje-Iveland pegmatites.
Oxygen isotope fractionations between quartz, K-feldspar, and garnet indicate equilibrium crystallization temperatures between approximately 850 and 600 E C for these Evje-Iveland pegmatite minerals. But other mineral pairs or triplets give higher or lower equilibrium isotope temperatures. For instance, the triplet quartz,
K-feldspar, and magnetite give temperatures between approximately 650 and 550 E C. A crystallization temperature ~ 600 E C would fit the observation that the quartz in the pegmatites always is of the β-variety [lowquartz] (Barth 1928) formed at ~ 600 E C, considering an unknown pressure of some 2 to 5 kbar. Hydrogenisotope fractionation between water released from fluid inclusions and coexisting muscovite and biotite give equilibrium temperatures of approximately 680 and 600 E C. The δD of this water has a range of -32 to -67‰ (for of -2.2 to -6.4‰ (Taylor & Friedrichsen 1983b).
The second, hydrothermal, stage of the pegmatite formation was looked upon by Taylor & Friedrichsen (1983a).
that there was closed-system oxygen-isotope exchange between entrapped aqueous fluid and pegmatite minerals during cooling. There is a question if the fluid responsible for the second stage of pegmatitic replacement (the cleavelandite stage) came from external sources; or if the fluid was the leftover from the first stage mineralizations, that now was in disequilibrium with earlier formed minerals. Taylor & Friedrichsen (1983a) pegmatite itself, and probably represents at least part of the water exsolved from the hydrous pegmatite melt.
Finally, Taylor & Friedrichsen (1983a) describe isotope analysis of minerals in a late breccia in the Landsverk
I mine, indicating influx of meteoric water at a very late stage, unrelated to the pegmatite petrogenesis. They suggest that this process matches the reported transition of amazonite to a brick-red K-feldspar in the same pegmatite. Taylor et al. (1960) interpreted this as a result of late-stage processes unrelated to magmatic differentiation. Assuming a temperature of 250 E C, Taylor & Friedrichsen (1983a) calculated a δ O of -16.0‰ for the water at this stage.
The Evje-Iveland amphibolite has been subjected to complete recrystallization; all its minerals have been formed during metamorphism (plagioclase and hornblende being the dominant minerals). Barth (1947) even introduced the rock-name evjite (named from Evje) for a eugranitic hornblende gabbro. The first direction of deformation was NE-SW; the second direction of deformation was E-W, judged from the directions of fold axes.
Barth (1947) was of the opinion that the pegmatites were formed as a result of intense granitization (partial anatexis) of the area. He presents calculations on how a metasomatic process, associated with regional amphibolitization of a noritic rock. The same mobile elements contribute in forming the pegmatites in available fractures in the area. We may add that transported incompatible elements form the rare minerals near the end of the pegmatitic stage. Hence these pegmatites are considered the results of a regional metamorphic-metasomatic process rather than a primary magmatic process.
Fluid-inclusion studies by Taylor & Friedrichsen (1983a) revealed saline C – O – H fluids containing liquid CO
2 and daughter crystals of possibly NaCl and an elongated anisotropic mineral (muscovite?) in the “magmatic stage” quartz. Recent fluid-inclusion studies by Larsen et al. (1998) found that the late magmatic volatile phase
2 2
2 2 2 2 formation of the core zone.
Chondrite-normalized REEs in pegmatite K-feldspar from Evje-Iveland show a symmetric trough broken by a moderate to strong Eu peak; the LREEs define a steeply falling slope, whereas the HREEs define an increasing slope; overall close to chondritic values (Larsen 2002). The distribution of REEs in K-feldspar in pegmatite-

N G F A bstracts and Proceedings, N o. 2, 2003
75 forming environments is buffered by various REE-bearing minerals forming throughout the solidification of a pegmatite. For Evje-Iveland, Larsen (2002) found that the LREEs and the middle REEs are being removed during the differentiation from the core-wall zone to the core zone, flattening the chondrite-normalized REE pattern. This is explained by the early crystallization of monazite-(Ce) and allanite-(Ce) being enriched in
LREEs, and by the later crystallization of Y- and HREE-enriching xenotime-(Y), euxenite-(Y), and fergusonite-
(Y) in the pegmatites (Bjørlykke 1935).
Larsen et al. (1997) and Larsen (2002) assumed an origin of the pegmatites from a primitive parental magma values, and the location of the most evolved pegmatites, having the highest Rb/Sr values, Larsen (2002) proposed that the pegmatite-forming melts/fluids propagated towards the S and SE from an origin somewhere in the N or
NW , along a major conduit along the western edge of the pegmatite field and, further south, branched out eastward. Trace elements in quartz showed that the concentration of Nb + Y was increasing toward the south, the suggested trend of differentiation. Larsen et al. (2000) simply claim, on the basis of the Rb/Sr-isotope data
(reviewed above), that the Høvringsvatn granite-monzonite igneous complex constituted the inferred parent magma of the granite pegmatites.
We will, however, assert that Barth’s (1947) partial anatexis model is not rejected by the new trace-element data nor the available Rb/Sr isotope data. Barth’s model involves a regional mobility and addition of K, possibly from a Rb-poor source (Pedersen 1980), resulting in the high K/Rb vs. low Rb for the “very primitive” granitic likely have been inherited from its parent magma. The Høvringsvatn monzonite-granite complex rocks have, enclosing the pegmatite-bearing amphibolites.
We will therefore, for the time being, conclude that the available Rb/Sr isotope data better support Barth’s partial anatexis and metasomatic model, involving a large-scale regional “granitization” process mobilizing certain elements now to be found in the pegmatites; rather than an ortho-magmatic model for the petrogenesis of the
Evje-Iveland granite pegmatites. Any systematic trace-element gradients in the area can be governed by different degrees of partial melting, wall-rock reactions, and a variety of other geochemical mechanisms.
The enrichment in Hf vs. Zr, and Ta vs. Nb, in the second, hydrothermal, cleavelandite stage of granitic pegmatites was noted by various researchers (e.g., Bjørlykke 1935; Levinson & Borup 1960a, 1960b), and has also been found elsewhere, e.g., in pegmatites in Mozambique (Correia Neves et al. 1974). Linnen (1998) conducted experiments to determine the solubility of columbite, tantalite, zircon, and hafnon in water-saturated granitic melts. He estimated that at 600 E C the granitic melt would be saturated in columbite but not tantalite.
However, tantalite saturation is predicted if the melts contained less F and Li. Therefore, the later genesis of tantalum mineralization may be explained by Ta being retained in the early melt because of its high F (± Li) concentration. Hence, tantalite crystallization will be delayed until F ± P ± Li minerals crystallize from the melt, considering the pegmatite a closed system. Linnen (1998) also found a higher solubility of hafnon than that of zircon in the water-saturated granitic melt at 600 E C, similar to the higher solubility of tantalite compared to columbite. These findings can explain why Zr/Hf and Nb/Ta ratios both decrease with fractionation in the granite pegmatite melt.
The following generalized paragenetic sequence for the “magmatic” stage of the complex granite pegmatites is modified after Bjørlykke (1935, 1937a), omitting the early granitic zone and the graphic granite zone, and omitting the late hydrothermal (cleavelandite) stage:

76
N G F A bstracts and Proceedings, N o. 2, 2003
MINERAL CORE-WALL ZONE CORE ZONE
Zirconium minerals ——
REE phosphate minerals ————
REE +(Nb,Ta) oxides ————————
(Fe,Mn)+(Nb,Ta) oxides ———
REE silicates ———————
Biotite —————————————————————
Muscovite —————
Thortveitite ——
Ilmenite ————————
Magnetite ————————
Gadolinite —
Allanite-(Ce) —
Beryl ——
Chrysoberyl ——
Spessartine —
Plagioclase ——
Microcline ———
Quartz ————
Bjørlykke (1935, 1937a) subdivided the pegmatites into Ca-rich and Ca-poor. He further subdivided the latter into four, based on their types of dark rare minerals: (1) thalenite – gadolinite type; (2) fergusonite type; (3) euxenite (samarskite) type; (4) columbite type. Each of these subtypes was further subdivided in three, based on increased Ti vs. Nb + Ta.
Iveland area (Schetelig 1911), there has been a certain interest in the geochemistry of scandium in this “scandium province”. Goldschmidt (1934) maintained that the scandium content of the thortveitite could not have come from the granite pegmatite magma, but rather from the gabbro-amphibolite wall-rock. His scandium analyses showed the reason for this, that felsic magmas in general have much less Sc than mafic rocks. Assimilation of mafic igneous rocks can supply Sc to the formation of thortveitite, from the breakdown of the original pyroxenes in the mafic igneous rocks. Bjørlykke (1937b) cites Goldschmidt’s Sc analyses (as a personal communication) on the amphibolite wall-rock next to one of the thortveitite-bearing pegmatites. Here Goldschmidt found that
Sc had been supplied to the pegmatite, by depleting the amphibolite wall-rock next to the pegmatite.
Oftedal (1943) reported Sc contents of 50 to 150 ppm in amphibole in amphibolite from Iveland. In normal granite pegmatite from Evje-Iveland he reported Sc contents of 50 to 1,000 ppm in biotite, and 5 to 1,000 ppm in muscovite. The Torvelona and Håverstad thortveitite-bearing occurrences had the highest Sc content in biotite and muscovite, respectively. Biotite from pegmatites without thortveitite had biotite with less than 100 ppm Sc.
One “ordinary green beryl” from the thortveitite-bearing pegmatite at Knipan was reported to contain some
10,000 ppm Sc (= 1% Sc); could there have been inclusions of thortveitite in this beryl? Two muscovite samples from the second stage (cleavelandite pegmatite) showed both less that 3 ppm Sc.
Neumann (1961) listed thortveitite occurrences in the Evje-Iveland area, and reported the scandium content of
176 Norwegian mineral samples. In an attempt to investigate the possibility of using beryl from Evje-Iveland as a scandium ore, he analysed 57 samples of beryl from this area, showing 10 to 1,000 ppm Sc. He found that beryl samples from thortveitite-bearing pegmatites generally were richer in Sc than elsewhere. Ilmenorutile was ilmenorutile from Iveland was recently described by Černý et al. (2000) and Černý & Chapman (2001).] One xenotime from Evje was reported to contain 10,000 ppm Sc. Davidites from Iveland contain 200 - 500 ppm Sc.
Ilmenite shows Sc contents of 30 to 1,000 ppm, with the highest values associated with thortveitite occurrences.

N G F A bstracts and Proceedings, N o. 2, 2003
77
Neumann (1961) therefore concluded that minerals from thortveitite-bearing pegmatites had higher Sc contents than the same minerals from pegmatites elsewhere. However, for the non-thortveitite-bearing pegmatites in the
Evje-Iveland area, Neumann tentatively concluded that there is no systematic difference between the Sc content of these and the Sc content of pegmatites in other areas in Norway. Neumann (1961) suggested that the conditions for the formation of thortveitite could be governed by lower than normal content of ferrous iron in
Although no primary minerals of tin have been found in the Evje-Iveland pegmatites, it is worth mentioning that thortveitite contains about 0.25% Sn; 0.01% Zr; and Fe, Mg, and Mn as minor elements (Oftedal 1969). This led
Sc may be mobilized by excess HF due to the formation of strong aqueous or melt fluoro complexes, e.g., of F-containing minerals, like early forming fluorite or biotite. If the aqueous- or melt-complexed Sc is losing its complexing agent, Sc can no longer stay in the solution or melt, and will have to crystallize together with a in the analogous site in muscovite. When the micas are saturated with Sc, and no other suitable crystal structures are available for Sc substitution, Sc will have to form a separate mineral. If other possible anion groups are not suitable (with respect to ionic radius and/or charge) or not available, Sc will form a separate Sc mineral like thortveitite with the ubiquitous silica (silicic acid) of the hydrous pegmatite melt. We may propose a possible simplified chemical reaction for this thortveitite-forming process (the aq designation here denotes a hydrous pegmatite melt; s denotes a solid):
2 Sc ( aq ) + 2 H SiO ( aq ) = Sc Si O ( s ) + H O + 6 H
+
A complete chemical reaction equation for the breakdown of the scandium hexafluoride complex, forming fluorite with available calcium, and thortveitite with available silicic acid, may look like this (to illustrate the removal of scandium’s complexing agent) :
2 [ScF ] ( aq ) + 6 Ca ( aq ) + 2 H SiO ( aq ) = Sc Si O ( s ) + 6 CaF ( s ) + H O + 6 H
+
The possibility of mica taking up fluorine, and for taking up scandium in the mica until saturation, is likely, because this model would also explain the observed high content of Sc in mica occurring in thortveitite-bearing pegmatites in the Evje-Iveland area. Further testing of the model can be done by analysing the F content of the micas throughout the paragenetic sequences from both thortveitite-bearing and non-thortveitite-bearing pegmatites in the Evje-Iveland area. The model’s principle can be modified to explain the formation of other Sc minerals, like for those occurring at Tørdal.
The Bjordammen plagioclase pegmatite in Bamble is an irregular lens 15 m long and 5 m wide, with strike 60 E and dip 70 E to SE, conformably surrounded by Precambrian gneisses. The pegmatite has the following minerals:
Plagioclase, quartz, biotite, minor hematite, amphibole, tourmaline, and rarely rutile and apatite. K-feldspar has never been found here. A major find of prismatic crystals of xenotime-(Y) several cm long was made in this pegmatite in the 1960s. The wall-rock to the NW is amphibolite, altered to biotite schist along the border. The wall-rock to the SE is a cordierite-anthophyllite-quartz schist (Neumann 1960b), whose cordierite contains 20 ppm Sc (Neumann 1961). No scandium minerals have been found at this locality.
This pegmatite is very rich in aventurine feldspar (“sunstone”), a red oligoclase with small (less than 0.2 mm), crystallographically oriented hematite lamellae (Andersen 1915; Copley & Gay 1978). The red colour of the feldspar is due to the smallest hematite inclusions (Neumann & Christie 1962). Andersen (1915) ascribed the formation of aventurine feldspar to unmixing of an originally homogeneous iron-rich feldspar. Divljan (1960)

78
N G F A bstracts and Proceedings, N o. 2, 2003 and Copley & Gay (1982) rather support the formation of aventurine feldspar through introduction of foreign material from an external source along planes of weakness (cleavage and composition planes) in normal plagioclase. Clouded red feldspars formed by precipitation of hematite is a common result of iron-bearing water
– rock interaction (Boone 1969). Cordierite in the wall-rock here has also been found to contain similar hematite lamellae.
Larvikite, a monzonitic rock, is the southernmost exposed plutonic rock of the Oslo Rift. It is also the most voluminous plutonic rock of the exposed rift, especially in the southern part of the O slo Region, where the the larvikite feldspars display a schiller-effect due to cryptoperthitic lamellae. This bluish schiller-effect makes the rock attractive as an ornamental and building stone, being extensively quarried in this area under trade names like “Labrador”, “Blue Pearl”, “Emerald Pearl”, etc. In these quarries a number of small nepheline-syenite pegmatites with rare minerals have been found (Andersen et al. 1996). These pegmatites resemble the nephelinesyenite pegmatites of the Langesundsfjord area (to the immediate SW of the larvikite SW border), made famous for their richness in extremely rare minerals (Brøgger 1890). The most famous of these pegmatites is located on the island of Låven in Langesundsfjord and is now protected by law. The pegmatites of the Langesundsfjord area, which is close to the western margin of the Oslo Rift, occur in ditroite, a medium-grained, gneissic nepheline syenite which shows signs of assimilation of bedrock.
Raade (1973) found through petrographic studies that the larvikite has both silica-saturated (quartz-bearing) and silica-undersaturated (nepheline-bearing) varieties. Petersen (1977) found through field work and structural analyses that the larvikite batholith comprises a complex of multiple (10 major) circular intrusions, whose centres have moved first toward W and then toward N (Fig. 5). Lardalite (a coarse-grained nepheline-syenite rock) and hedrumite (foyaite) are the major occurrences of silica-undersaturated plutonic rocks in the Oslo Rift (IX and
X, respectively, in Fig. 5), and constitute an important silica-undersaturated end point for the fractional crystallization of this branch of the Oslo Rift magmatism (Fig. 1). This very end point is also reached, on a minor scale, in the nepheline-syenite pegmatites occurring in conjunction with the larvikite.
Fig. 5. The larvikite-lardalite batholith complex of the southern Oslo Region; redrawn after Raade (1973) and Petersen
(1977).

N G F A bstracts and Proceedings, N o. 2, 2003
79
The results of U-Pb dating of zircon and baddeleyite in the plutons and pegmatites of the Larvik ring complex were published by Dahlgren et al. (1998). It was shown that the plutons were emplaced over a 5 to 6 m.y. long period, between about 297 and 292 Ma. These results are different from the much younger Rb-Sr ages previously reported from these plutons which gave 281 ± 4 Ma (Sundvoll & Larsen 1990) and 277 ± 3 (Sundvoll et al.
1990). An important find was that the pegmatites and their surrounding plutons give identical U-Pb ages within the analytical error, supporting the view of the pegmatites being a late-stage differentiation product of the surrounding magmatic rock.
There are very small chemical differences between the larvikite intrusion zones in terms of major and trace elements (Neumann et al. 1977, Neumann 1980). The most apparent change through the series is that the lardalite becomes gradually richer in Ti, Fe, Mg, and P towards its centre (X in Fig. 5). Both crystal accumulation and crystal fractionation have been important processes in the development of the larvikite-lardalite batholith.
Generally this rock series is highly enriched in large ionic lithophile elements (LILE). These rocks have low and to have a mantle origin with none or very minor crustal contamination (Rasmussen et al. 1988). Raade (1973,
1978) found a fairly constant Th/U ratio of 3.5 to 3.7 for this rocks series, in spite of highly variable contents of Th and U. This also points to an uncontaminated mantle origin, which was also found from oxygen-isotope work (Segalstad & Ohmoto 1980).
Because of the very small chemical differences between different larvikite intrusions, we decided to try using multivariate data techniques to reveal any trace element variation and their possible causes. The parametric statistics computer program CORFAN (Ondrick & Srivastava 1970) was used for this purpose. Factor analysis on 15 log-transformed, normalized trace elements in 42 samples of larvikite from all parts of the southern larvikite batholith identified three significant factors explaining 84.3% of all total variance (variance and covariance) of the dataset. Varimax rotation around the three factors gave high factor loadings for La, Ce, Sm, and Tb for the first (most significant) factor; for Eu, Sr, and Ba for the second factor; and for Rb, Cs, Hf, Ta, Th, and U for the third factor. Yb and Lu did not give high factor loadings for the rotated matrix. The three factors were identified as representing apatite, feldspar, and zircon, respectively, from comparison with partitioncoefficient data. These three minerals seem to have the major responsibility for the trace-element variation in the batholith, governing what elements are available for rare-mineral formation at the nepheline-syenite pegmatite stage. The rare minerals here comprise varying compositions of Ti, Zr, REE, F, Nb, Be, B, etc., with
Ca and alkalis.
The nepheline-syenite pegmatites occurring within the larvikite are most likely the true closed-system end product of the larvikite differentiation. The nepheline-syenite pegmatites which intruded into nearby basalts and sediments, however, show the highest variety of extremely rare minerals. Brøgger (1890) claimed that the mineralogy of the Langesundsfjord pegmatites was strongly dependent on resorption of the basic constituents of the “Augitporphyrit” (= the basaltic rocks) during the injection of the nepheline-syenite pegmatite magma.
A resorption of wall-rock constituents has evidently taken place in some of these cases. In the Langesundsfjord area the nepheline-syenite pegmatites penetrating the basaltic rocks change to quartz-bearing granite pegmatites when penetrating the Ringerike sandstone (Brøgger 1890, Dons 1969). Segalstad & Larsen (1978a) presented a further discussion about this, and showed how the wall-rocks had obtained an La-, Nd-, and Ce-enriched gradient from the pegmatite magma. When the pegmatite magma intruded into wall-rocks outside the larvikite, a two-way exchange of chemical elements evidently took place, contributing to the peculiar mineralogy of these pegmatites.
Sc contents in minerals from the nepheline-syenite pegmatites in the Langesundsfjord area were published by
Neumann (1961): each of the minerals zircon, eudialyte (eucolite), hiortdahlite, catapleiite, rosenbushite, mosandrite (johnstrupite), and wöhlerite has 10 ppm Sc. One sample of larvikite from Tjølling was found to have
150 ppm Sc (Oftedal 1943). No scandium minerals are known in these pegmatites.
About 180 different minerals have been found in the nepheline-syenite pegmatites of the Larvik and
Langesundsfjord districts (Andersen et al. 1996). The following minerals were described for the first time from these pegmatites, in chronological order (Raade 1996): pyrochlore, thorite, aegirine, leucophanite, mosandrite, wöhlerite, catapleiite, tritomite-(Ce), meliphanite, astrophyllite, homilite, cappelenite-(Y), låvenite, melano-

80
N G F A bstracts and Proceedings, N o. 2, 2003 cerite-(Ce), nordenskiöldine, rosenbuschite, eudidymite, hambergite, hiortdahlite, gadolinite-(Ce), chiavennite, and tvedalite. Of particular interest is the wide range of Be-bearing minerals in these pegmatites, formed over a wide temperature range (in alphabetical order): barylite, bavenite, behoite, berborite, bertrandite, bromellite, chiavennite, epididiymite, eudidymite, phenakite, gadolinite-(Ce), genthelvite, hambergite, helvite, hingganite-(Y), chrysoberyl, leifite, leucophanite, meliphanite, milarite, and tvedalite. This is nearly one fourth of all known Be minerals.
The composition of the syenite pegmatites varies roughly according to the composition of the hosting circular zone of larvikite (Fig. 5): in the north-east, the pegmatites are quartz-bearing, and in the south and west, they are nepheline-bearing. The central and latest intrusion of foyaite has highly agpaitic pegmatites with parakeldyshite and other sodium-rich minerals (Raade & Mladeck 1977).
We would like to point out some main geochemical and mineralogical differences between the Norwegian granite pegmatites and nepheline-syenite pegmatites: the rare-element granite pegmatites typically carry oxides of Nb, Ta, U, Y and heavy REEs, and are in special cases enriched in Li, Sn, Sc, B, P, or As; the nephelinesyenite pegmatites are above all characterized by sodium-rich minerals of Zr and Ti and are enriched in light
REEs. Important elements common for both pegmatite types are Be, B, and F.
Åmli, R. 1977: Carbonatites, a possible source of scandium as indicated by Sc mineralization in the Fen Peralkaline
Complex, southern Norway. Economic Geology 72 , 855-869.
Andersen, F., Berge, S.A. & Burvald, I. 1996: Die Mineralien des Langesundsfjords und des umgebenden Larvikit-Gebietes,
Oslo-Region, Norwegen. Mineralien-Welt 7(4), 21-100.
Andersen, O. 1915: On aventurine feldspar. American Journal of Science 4th Series 40, 351-399.
Andersen, T. 1983: Iron ores in the Fen Central Complex, Telemark (S. Norway): petrography, chemical evolution and conditions of equilibrium. Norsk Geologisk Tidsskrift 63, 73-82.
Andersen, T. 1984: Secondary processes in carbonatites: petrology of “rødberg” (hematite-calcite-dolomite carbonatite) in the Fen Central Complex, Telemark (South Norway). Lithos 17, 227-245.
Andersen, T. 1986: Magmatic fluids in the Fen Carbonatite Complex, S.E. Norway: evidence of mid-crustal fractionation from solid and fluid inclusions in apatite. Contributions to Mineralogy and Petrology 93, 491-503.
Andersen, T. 1987: Petrogenesis and G eochemistry of Carbonatites: the Fen Complex, SE Norway.
Dissertation for the degree of Doctor Philosophiae, University of Oslo.
Anthony, E.Y., Segalstad, T.V. & Neumann, E.-R. 1989: An unusual mantle source region for nephelinites from the Oslo
Rift, Norway. Geochimica et Cosmochimica Acta 53, 1067-1076.
Bailey, S.W. & Christie, O.H.J. 1978: Three-layer monoclinic lepidolite from Tørdal, Norway. American Mineralogist 63 ,
203-204.
Barth, T.F.W. 1928: Zur Genese der Pegmatite im Urgebirge, II. Chemie der Erde 4, 95-136.
Barth, T.F.W. 1931: Feltspat III. Forekomster i Iveland og Vegusdal i Aust-Agder og i flere herreder i Vest-Agder. Norges
Geologiske Undersøkelse 128 b, 111-154.
Barth, T.F.W. 1944: Studies on the igneous rock complex of the Oslo Region, II. Systematic petrography of the plutonic rocks. Det Norske Videnskaps-Akademi i Oslo, I. Matematisk-Naturvitenskapelig Klasse, No. 9, 104 pp.
Barth, T.F.W . 1947: The nickeliferous Iveland-Evje amphibolite and its relation. Norges Geologiske Undersøkelse 168 a,
71 pp. + 1 map.
Barth, T.F.W. & Ramberg, I.B. 1966: The Fen circular complex. In Tuttle, O.F. & Gittins, J. (eds.): Carbonatites.
Interscience Publishers (John Wiley & Sons), 225-257.
Bergstøl, S. & Juve, G. 1988: Scandian ixiolite, pyrochlore and bazzite in granite pegmatite in Tørdal, Telemark, Norway.
A contribution to the mineralogy and geochemistry of scandium and tin. Mineralogy and Petrology 38 , 229-243.
Bergstøl, S. & Svinndal, S. 1960: The carbonatite and peralkaline rocks of the Fen area: Petrology. Norges Geologiske
Undersøkelse 208, 99-105.
Bergstøl, S., Jensen, B.B. & Neumann, H. 1977: Tveitite, a new calcium yttrium fluoride. Lithos 10 , 81-87.
Bergstøl, S., Juve, G. & Wilberg, R. 1985: Fordeling av tinn og andre metaller i Nissedal-området. Malmgeologisk symposium "Nye malmtyper i Norge".
Bergverkenes Landssammenslutnings Industrigruppe, Bergforskningen, 101-109.
Bevan, D.J.M. & Lawton, S.E. 1986: Cuboctahedral anion clusters in fluorite-related superstructures: geometrical calculations. , 55-58.
Bevan, D .J.M., Strähle, J. & Greis, O. 1982: The crystal structure of tveitite, an ordered yttrofluorite mineral. Journal of
Solid State Chemistry 44 , 75-81.

N G F A bstracts and Proceedings, N o. 2, 2003
81
Bjørlykke, H. 1935: The mineral paragenesis and classification of the granitic pegmatites of Iveland, Setesdal, South
Norway. Norsk Geologisk Tidsskrift 14, 211-311 + 1 plate + 1 map.
Bjørlykke, H. 1937a: The granite pegmatites of southern Norway. American Mineralogist 22, 241-255.
Bjørlykke, H. 1937b: Litt om granittiske pegmatittganger. Norsk Geologisk Tidsskrift 16, 299-303.
Boone, G.M. 1969: Origin of clouded red feldspars: petrologic contrasts in a granitic porphyry intrusion. American Journal of Science 267, 633-668.
Boudin, A. & Deutsch, S. 1970: Geochronology; recent development in the lutetium-176 / hafnium-176 dating method.
Science 168, 1219-1220.
Bowen, N.L. 1924: The Fen Area in Telemark, Norway. American Journal of Science 8(43), 1-11.
Bowen, N.L. 1926: The carbonate rocks of the Fen Area in Norway. American Journal of Science 12(72), 499-502 + 3
Plates.
Bowen, N.L. 1928: The Evolution of the Igneous Rocks.
Princeton University Press, 334 pp.
Brøgger, W.C. 1890: Die Mineralien der Syenitpegmatitgänge der südnorwegischen Augit- und Nephelinsyenite. Zeitschrift für Krystallographie 16, 1-898.
Brøgger, W.C. 1921: Die Eruptivgesteine des Kristianiagebietes. IV. Das Fengebiet in Telemark, Norwegen.
Videnskapsselskapets Skrifter (Kristiania). I. Matematisk-Naturvidenskapelig Klasse 1920, No. 9, 408 pp. + 30 plates
+ 2 maps.
Buch, L. von 1810: Reise durch Norwegen und Lappland.
G.C. Nauck, Berlin; Vol. I, 486 pp. & Vol. II, 406 pp.
Bugge, A. 1963: Norges molybdenforekomster. Norges Geologiske Undersøkelse 217, 134 pp.
Bugge, C. 1917: Kongsbergfeltets geologi. Norges Geologiske Undersøkelse 82, 272 pp. + 11 plates (1 map + 7 plates in envelope).
Bugge, J.A.W. 1978: Norway. In Bowie, S.H.U., Kvalheim, A. & Haslam, H.W. (eds.): Mineral Deposits of Europe. Vol.
1: Northwest Europe.
The Institution of Mining and Metallurgy & The Mineralogical Society, 199-249.
Burnham, C.W. & Ohmoto, H. 1980: Late-stage processes of felsic volcanism. Mining Geology [Japan] Special Issue 8,
1-11.
Černý, P. 1992: Geochemical and petrogenetic features of mineralization in rare-element granitic pegmatites in the light of current research. Applied Geochemistry 7 , 393-416.
Č erný, P. & Chapman, R. 2001: Exsolution and breakdown of scandian and tungstenian Nb-Ta-Ti-Fe-Mn phases in niobian rutile. Canadian Mineralogist 39 , 93-101.
Černý, P. & Ercit, T.S. 1985: Some recent advances in the mineralogy and geochemistry of Nb and Ta in rare-element granitic pegmatites. Bulletin de Minéralogie 108 , 499-532.
Č erný, P., Chapman, R. & Masau, M. 2000: Two-stage exsolution of a titanian (Sc,Fe 3+ )(Nb,Ta)O phase in niobian rutile
4 from southern Norway. Canadian Mineralogist 38 , 907-913.
Copley, P.A. & Gay, P. 1978: A scanning electron microscope investigation of some Norwegian aventurine feldspars. Norsk
Geologisk Tidsskrift 58, 93-95.
Copley, P.A. & Gay, P. 1982: The heat treatment of some Norwegian aventurized feldspars. Mineralogical Magazine 45,
107-110.
Correia Neves, J.M., Lopes Nunes, J.E. & Sahama, T.G. 1974: High hafnium members of the zircon-hafnon series from the granite pegmatites of Zambézia, Mozambique. Contributions to Mineralogy and Petrology 48, 73-80.
Dahlgren, S. 1983: Naturlig radioaktivitet i berggrunnen. Gammastrålingskart. Fensfeltet, Telemark, m 1:10.000. Prosjekt
Temakart, Telemark, Fylkeskartkontoret i Telemark.
Dahlgren, S. 1994: Late Proterozoic and Carboniferous ultramafic magmatism of carbonatitic affinity in southern Norway.
Lithos 31, 141-154.
Dahlgren, S., Corfu, F. & Heaman, L. 1998: D atering av plutoner og pegmatitter i Larvik pluton-kompleks, sydlige Oslo
Graben, ved hjelp av U-Pb isotoper i zirkon og baddeleyitt. Norsk Bergverksmuseum Skrift 14 , 32-39.
Dawson, J.B. 1962: The geology of Oldoinyo Lengai. Bulletin of Volcanology 24, 349-387.
Divljan, S. 1960: The results of field and laboratory studies of aventurine plagioclases from some Norwegian pegmatites.
In Neumann, H. & Neuvonen, K. (eds.): Report of the Twenty-First Session Norden, International Geological Congress,
Part XVII, Proceedings of Section 17. Minerals and Genesis of Pegmatites.
Berlingske, Copenhagen, 94-101.
Dons, J.A. 1969: Interrupted creation of nepheline syenite pegmatite dikes in the Langesundsfjord area, S. Norway. Norsk
Geologisk Tidsskrift 49, 145-157.
Dypvik, H. 1977: Mineralogical and geochemical studies of lower Paleozoic rocks from the Trondheim and Oslo regions,
Norway. Norsk Geologisk Tidsskrift 57, 205-241.
Eckermann, H. von 1942: Et preliminärt meddelande om nya forskningsrön inom Alnö alkalina område. Geologiska
Föreningen i Stockholm Förhandlingar 64, 399-455.
Eckermann, H. von 1948: The alkaline district of Alnö Island (Alnö alkalina område). Sveriges Geologiska Undersökning,
Ser. Ca 36, 176 pp. + 56 plates + 4 maps.
Eldjarn, 2002: Seltene Scandium-Mineralien von Tørdal, Telemark, Norwegen. Mineralien-Welt 13(2), 14-22.
Ferraris, G., Gula, A., Ivaldi, G., N espolo, M . & Raade, G. 2001: Crystal structure of kristiansenite: a case of class IIB twinning by metric merohedry. Zeitschrift für Kristallographie 216 , 442-448.
Frigstad, O.F. 1999: Amazonittpegmatitter i Iveland-Evje. Norsk Bergverksmuseum Skrift 15, 60-73.

82
N G F A bstracts and Proceedings, N o. 2, 2003
Frondel, C. 1968: Crystal chemistry of scandium as a trace element in minerals. Zeitschrift für Kristallographie 127 , 121-
138.
Frøyland, M. & Segalstad, T.V. 1992: Geochemistry of lead-zinc sulfide ore deposits in the Fiskum area, southern Norway.
20th Geological Winter Meeting, Reykjavik, Iceland, Abstracts, 44.
Goldschmidt, V.M. 1934: Drei Vorträge über Geochemie. Geologiska Föreningen i Stockholm Förhandlingar 56, 385-427.
Gramaccioli, C.M., Diella, V. & Demartin, F. 2000: The formation of scandium minerals as an example of the role of complexes in the geochemistry of the rare earths and HFS elements. European Journal of Mineralogy 12, 795-808.
Greis, O. 1978: Pulverröntgenographische Untersuchungen und Einkristall – Elektronendiffraktion an Tveitit
Ca
12+ δ
(Y,SE) F
44-δ
Revue de Chimie Minérale 15 , 481-493.
Griffin, W .L. 1973: Lherzolite nodules from the Fen Alkaline Complex, Norway. Contributions to Mineralogy and
Petrology 38, 135-146.
Griffin, W.L. & Taylor, P.N. 1975: The Fen damkjernite: petrology of a “central-complex kimberlite”. Physics and
Chemistry of the Earth 9, 163-177.
Hawthorne, F.C. 2002: The use of end-member charge-arrangements in defining new mineral species and heterovalent substitutions in complex minerals. Canadian Mineralogist 40 , 699-710.
Heier, K.S. & Compston, W. 1969: Rb-Sr isotopic studies of the plutonic rocks of the Oslo Region. Lithos 2, 133-145.
Hestmark, G. 1999: Vitenskap og nasjon. Waldemar Christopher Brøgger 1851-1905.
Aschehoug, Oslo, 862 pp.
Hofmeister, A.M. & Rossman, G.R. 1985: A spectroscopic study of irradiation coloring of amazonite: structurally hydrous,
Pb-bearing feldspar. American Mineralogist 70, 794-804.
Högbom, A.E. 1895: Über das Nephelinsyenitgebiet auf den Insel Alnö. Geologiska Föreningen i Stockholm Förhandlingar
17, 100-160 + 1 figure + 1 map; + 214-256.
Ihlen, P.B. & Nordrum, F.S. 1986: The Kongsberg mining area - a road log. Sveriges Geologiska Undersökning Ser. Ca 59,
33-38.
Ihlen, P.B. & Vokes, F.M. 1978: Metallogeny. In Dons, J.A. & Larsen, B.T. (eds.): The Oslo Paleorift. A review and guide to excursions. Norges Geologiske Undersøkelse 337, 75-90.
Ineson, P.R., M itchell, J.G. & Vokes, V.M. 1975: K-Ar dating of epigenetic mineral deposits: an investigation of the
Permian metallogenic province of the Oslo Region, southern Norway. Economic Geology 70, 1426-1436.
Jacobsen, S.B. & Wasserburg, G.J. 1978: Nd and Sr isotop ic study of the Permian Oslo Rift. In Zartman, R.E. (ed.): Short
Papers of the Fourth International Conference, Geochronology, Cosmochronology, Isotope Geology, 1978. U.S.
Geological Survey, Open-File Report 78-701, 194-196.
Johansen, H. 1985: Geokjemi av hydrotermale prosesser i de sølvførende kalkspatganger på Kongsberg.
Unpublished Cand.
Scient. Thesis, University of Oslo , 239 pp.
of amazonite. Physics and Chemistry of Minerals 25, 229-233.
Juve, G. & Bergstøl, S. 1990: Caesian bazzite in granite pegmatite in Tørdal, Telemark, Norway. Mineralogy and Petrology
43, 131-136.
Juve, G. & Bergstøl, S. 1997: Granittpegmatitter i Tørdal, Telemark. Norsk Bergverksmuseum Skrift 12 , 56-57.
Kristiansen, R. 1998: Høydalen Li-pegmatitt, Tørdal, Telemark. Norsk Bergverksmuseum Skrift 14 , 17-28.
Larsen, R.B. 2002: The distribution of rare-earth elements in K-feldspar as an indicator of petrogenetic processes in granitic pegmatites: examples from two pegmatite fields in southern Norway. Canadian Mineralogist 40, 137-151.
Larsen, R.B., Polvé, M., Juve, G. 2000: Granite pegmatite quartz from Evje-Iveland: trace element chemistry and implications for the formation of high-purity quartz. Norges Geologiske Undersøkelse Bulletin 436, 57-65.
Larsen, R.B., Polvé, M., Juve, G. & Poitrasson, F. 1997: Composition of volatiles and structural admixtures in quartz in granitic pegmatites, Evje-Iveland, South Norway. Norges Geologiske Undersøkelse Bulletin 433, 38-39.
Levinson, A.A. & Borup, R.A. 1960a: High hafnium zircon from Norway. American Mineralogist 45, 562-565.
Levinson, A.A. & Borup, R.A. 1960b: N ew data on the hafnium, zirconium and yttrium content of thortveitite. American
Mineralogist 45, 712-715.
Linnen, R.L. 1998: The solubility of Nb-Ta-Zr-Hf-W in granitic melts with Li and Li + F: constraints for mineralization in rare metal granites and pegmatites. Economic Geology 93, 1013-1025.
Meert, J.G., Torsvik, T.H., Eide, E.A. & Dahlgren, S. 1998: Tectonic significance of the Fen Province, S. Norway: constraints from geochronology and paleomagnetism. Journal of Geology 106, 553-564.
Mitchell, R.H. 1967: The Precambrian rocks of the Telemark Area in south-central Norway, V. The Nissedal supracrustal series. Norsk Geologisk Tidsskrift 47, 295-332.
Mitchell, R.H. & Brunfelt, A.O. 1974: Scandium, cobalt and iron geochemistry of the Fen Alkaline Complex, Norway.
Earth and Planetary Science Letters 23, 189-192.
Mitchell, R.H. & Brunfelt, A.O. 1975: Rare earth element geochemistry of the Fen Alkaline Complex, Norway.
Contributions to Mineralogy and Petrology 52, 247-259.
Mitchell, R.H. & Crockett, J.H. 1972: Isotopic composition in rocks of the Fen Alkaline Complex, South Norway. Journal of Petrology 13, 83-98.
Neumann, E.-R. 1980: Petrogenesis of the O slo Region larvikites and associated rocks. Journal of Petrology 21, 499-531.
Neumann, E.-R., Brunfelt, A.O. & Finstad, K.G. 1977: Rare earth elements in some igneous rocks in the Oslo Rift, Norway.

N G F A bstracts and Proceedings, N o. 2, 2003
83
Lithos 10, 311-319.
Neumann, E.-R., Sundvoll, B. & Ø verli, P.E. 1990: A mildly depleted upper mantle beneath southeast Norway: evidence from basalts in the Permo-Carboniferous Oslo Rift. Tectonophysics 178, 89-107.
Oslo rift: derivation from initial melts formed by progressive partial melting of an enriched mantle source. Lithos 61 ,
21-53.
Neumann, H. 1944: Silver deposits at Kongsberg. Norges Geologiske Undersøkelse 162, 133 pp.
Neumann, H. 1960a: Apparent ages of Norwegian rocks and minerals. Norsk Geologisk Tidsskrift 40 , 173-191.
Neumann, H. 1960b: Mineral occurrences in southern Norway. Norges Geologiske Undersøkelse 212 o, 18 pp.
Neumann, H. 1961: The scandium content of some Norwegian minerals and the formation of thortveitite, a reconnaissance survey. Norsk Geologisk Tidsskrift 41, 197-210.
Neumann, H. 1985: Norges mineraler. Norges Geologiske Undersøkelse Skrifter 68, 278 pp.
Neumann, H. & Christie, O.H.J. 1962: Observations on plagioclase aventurines from southern Norway. Norsk G eologisk
Tidsskrift 42, 389-393.
O’Brien, T.D. 1966: Scandium and compounds. In Clark, G.L. & Hawley, G.G. (eds.): The Encyclopedia of Chemistry, 2nd ed.
Reinhold Publ. Co., New York, 957-958.
Oftedal, I. 1942: Lepidolit- og tinnstensførende pegmatitt i Tørdal, Telemark. Norsk Geologisk Tidsskrift 47, 295-332.
Oftedal, I. 1943: Scandium in biotite as a geologic thermometer. Norsk Geologisk Tidsskrift 23, 202-213.
Oftedal, I. 1954: Some observations on the regional distribution of lead in south Norwegian granitic rocks. Norsk Geologisk
Tidsskrift 33 , 153-161.
Oftedal, I. 1956: Contribution to the geochemistry of granite pegmatite. Norsk Geologisk Tidsskrift 36 , 141-150.
Oftedal, I. 1957: Heating experiments on amazonite. Mineralogical Magazine 31, 417-419.
Oftedal, I. 1969: On minor elements in thortveitite. Norsk Geologisk Tidsskrift 49, 77-79.
Oftedal, I. 1970: Lithium contents of Norwegian beryls. Norsk Geologisk Tidsskrift 50, 245-247.
Ondrick, C.W. & Srivastava, G.S. 1970: CORFAN – FORTRAN IV computer program for correlation, factor analysis (Rand Q-mode) and varimax rotation. Kansas Geological Survey Computer Contribution 42, 92 pp.
Pedersen, S. 1973: Age determinations from the Iveland-Evje Area, Aust-Agder. Norges Geologiske Undersøkelse Bulletin
300, 33-39.
Pedersen, S. 1975: Intrusive rocks of the northern Iveland-Evje area, Aust-Agder. Norges Geologiske Undersøkelse Bulletin
322, 1-11.
Pedersen, S. 1980: Rb/Sr age determinations on late Proterozoic granitoids from the Evje area, South Norway. Bulletin of the Geological Society of Denmark 29, 129-143.
Petersen, J.S. 1977: Structure of the larvikite-lardalite complex, Oslo-Region, N orway, and its evolution. Geologische
Rundschau 67, 330-342.
microcline. American Mineralogist 78, 500-510.
Raade, G. 1973: Distribution of radioactive elements in the plutonic rocks of the Oslo Region, Norway.
Unpublished Cand.
Real. Thesis, University of Oslo, 162 pp.
Raade, G. 1978: Distribution of Th, U, K in the plutonic rocks of the Oslo Region, Norway. In Neumann, E.-R. & Ramberg,
I.B. (eds.): Petrology and Geochemistry of Continental Rifts.
D. Reidel Publ. Co., Dordrecht, 185-192.
Raade, G. 1996: Minerals originally described from Norway. Including notes on type material. Norsk Bergverksmuseum
Skrift 11 , 107 pp. (+ 7 Plates).
Raade, G. & Brastad, K. 1993: Kamphaugite-(Y), a new hydrous Ca-(Y,REE)-carbonate mineral. European Journal of
Mineralogy 5, 679-683.
Raade, G. & Erambert, M. 1999: An intergrowth of scandiobabingtonite and cascandite from the Heftetjern granite pegmatite, Norway. Neues Jahrbuch für Mineralogie Monatshefte , 545-550.
Raade, G. & Kristiansen, R. 2000a: Mineralogy and geochemistry of the Heftetjern granite pegmatite, Tørdal: a progress report. Norsk Bergverksmuseum Skrift 17 , 19-25.
Raade, G. & Kristiansen, R. 2002b: Scandium enrichment in the Heftetjern granite pegmatite, Telemark, Norway. Program and Abstracts Volume, 4th International Conference Mineralogy and Museums (Melbourne, Australia), 83.
Raade, G. & Mladeck, M.H. 1977: Parakeldyshite from Norway. Canadian Mineralogist 15 , 102-107.
Raade, G., Sæbø, P. C., Austrheim, H. & Kristiansen, R. 1993: Kuliokite-(Y) and its alteration products kainosite-(Y) and kamphaugite-(Y) from granite pegmatite in Tørdal, Norway. European Journal of Mineralogy 5 , 691-698.
Raade, G., Ferraris, G., Gula, A., Ivaldi, G. & Bernhard, F. 2002: Kristiansenite, a new calcium-scandium-tin sorosilicate from granite pegmatite in Tørdal, Telemark, Norway. Mineralogy and Petrology 75, 89-99.
Ramberg, I.B. 1973: Gravity studies of the Fen Complex, Norway, and their petrological significance. Contributions to
Mineralogy and Petrology 38, 115-134.
Ramberg, I.B. 1976: Gravity interpretation of the Oslo Graben and associated igneous rocks. Norges Geologiske
Undersøkelse Bulletin 325, 194 pp.
Ramberg, I.B. & Barth, T.F.W. 1966: Eocambrian volcanism in southern Norway. Norsk Geologisk Tidsskrift 46, 219-236.
Ramberg, I.B. & Larsen, B.T. 1978: Tectonomagmatic evolution. In Dons, J.A. & Larsen, B.T. (eds.): The Oslo Paleorift.

84
N G F A bstracts and Proceedings, N o. 2, 2003
A review and guide to excursions. Norges Geologiske Undersøkelse 337, 55-73.
Rasmussen, E., Neumann, E.-R., Andersen, T., Sundvoll, B., Fjerdingstad, V. & Stabel, A. 1988: Petrogenetic processes associated with intermediate and silicic magmatism in the Oslo Rift, south-east Norway. Mineralogical Magazine 52,
293-307.
Rule, A.C., Bailey, S.W., Livi, K.J.T. & Veblen, D.R. 1987: Complex stacking sequences in a lepidolite from Tørdal,
Norway. American Mineralogist 72 , 1163-1169.
Sæther, E. 1957: The alkaline rock province of the Fen Area in southern Norway. Det Kongelige Norske Videnskabers
Selskabs Skrifter 1957, No. 1. (Trondheim), 150 pp. + 2 maps.
Scherer, E., Münker, C. & Mezger, K. 2001: Calibration of the lutetium-hafnium clock. Science 293, 683-687.
Schetelig, J. 1911: Ueber Thortveitit, ein neues Mineral. Centralblatt für Mineralogie, Geologie und Paläontologie, 721-
726.
In Neumann, E.-R. & Ramberg, I.B. (eds.): Petrology and Geochemistry of Continental Rifts.
D. Reidel Publ. Co.,
Dordrecht, 209-216.
Segalstad, T.V. 1984: An unusual titanium-rich oxide mineral from Oslo, Norway. American Mineralogist 69 , 388-390.
Segalstad, T.V. 1996: The ore mineralization at the Kongsberg silver mines, Norway. In Jamtveit, B. (ed.): Fractures, Fluid
Flow and Transport in Fractures, 10th Kongsberg Seminar, 8-10 May 1996, Kongsberg, Norway.
Department of
Geology, University of Oslo, Abstracts and Program, 6-7.
[Available at: http://www.fys.uio.no/faststoff/sup/Meetings/AbsKongsberg96/index.html].
Segalstad, T.V. 2000: Native silver from Kongsberg. In Roaldset, E. & Sjulsen, S.-E. (eds.): Highlights - Selected
Attractions.
Natural History Museums and Botanical Garden, University of Oslo, 34-39.
Segalstad, T.V. 2001: Termokjemisk modellering av Kongsberg sølvforekomst. GEONYTT, Norsk Geologisk Forenings
XVII Vinterkonferanse, Oslo, 8-10 Januar 2001, Program & Sammendrag, 97.
Segalstad, T.V. & Eggleston, T.L. 1990: Petrogenesis of tin-bearing complex pegmatites in Southern Norway. Abstracts with Program, 8th IAGOD (International Association on the Genesis of Ore Deposits) Symposium, Ottawa, Canada,
A170-A171.
Segalstad, T.V. & Eggleston, T.L. 1993a: Pegmatittene i Tørdal, Telemark. Stein 20(3), 190-195.
Segalstad, T.V. & Eggleston, T.L. 1993b: Pressure / temperature relations in complex granite pegmatites. Gordon Research
Conference on Geochemistry of Hydrothermal Ore Deposits, Tilton, New Hampshire, 2 pp.
Segalstad, T.V. & Larsen, A.O. 1978a: Gadolinite-(Ce) from Skien, southwestern Oslo Region, Norway. American
Mineralogist 63, 188-195.
Segalstad, T.V. & Larsen, A.O. 1978b: Chevkinite and perrierite from the Oslo Region, Norway. American Mineralogist
63, 499-505.
Segalstad, T. V. & Ohmoto, H. 1980: Petrogenetic implications of oxygen isotope studies in the Oslo Rift, Norway. EOS
61, 410.
Segalstad, T. V. & Ohmoto, H. 1986: Magmatic and crustal waters in hydrothermal solutions associated with ore deposition in the Permian Oslo Rift, Norway. Terra Cognita 6, 555.
Segalstad, T.V. & Ohmoto, H. 1990: Geochemistry of hydrothermal ore forming processes in the Oslo Rift, Norway. 8th
IAGOD Symposium 1990, Ottawa, Canada, Abstracts with Program, Addendum, 1.
Segalstad, T. V. & Telstø, L. 2002: Mineraldannelsen i Konnerudkollen gruver. Norsk Bergverksmuseum Skrift 20, 35-49.
Segalstad, T.V., Johansen, H. & Ohmoto, H. 1986: Geochemistry of hydrothermal processes at the Kongsberg Silver
Deposit, Southern Norway. Terra Cognita 6, 511.
Shettel, D.L., Jr. 1978: Experimental determination of oxygen isotopic fractionation between H O and hydrous silicate melts.
2
Unpublished Ph. D. Thesis, Pennsylvania State University, 116 pp.
Stockmarr, P. 1994: A description of pegmatites at Åvesland and Evje, South Norway.
Unpublished M. Sc. Thesis,
University of Copenhagen, Denmark [referenced in Larsen 2002].
Sundvoll, B. 1994: The rhomb-porphyry dyke intrusions in the Oslo Rift: aspects of tectonomagmatic evolution. In Sundvoll,
B.: Magmatic and Tectonic Aspects of Rift Evolution: the Oslo (Paleo-) Rift.
Dissertation for the Degree of Doctor
Philosophiae, 29 pp. + 4 figures.
Sundvoll, B. & Larsen, B.T. 1990: Rb-Sr isotope systematics in the magmatic rocks of the O slo Rift. Norges Geologiske
Undersøkelse Bulletin 418, 27-46.
Sundvoll, B., Neumann, E.-R., Larsen, B.T. & Tuen, E. 1990: Age relations among Oslo Rift magmatic rocks: implications for tectonic and magmatic modelling. Tectonophysics 178 , 67-87.
Sverdrup, T.L. 1968: Contributions to the Mineralogy of Norway. No. 37. Y ttrofluorite – yttrocerite – cerfluorite in
Norwegian pegmatites. Norsk Geologisk Tidsskrift 48 , 245-252.
Sverdrup, T.L., Sæbø, P.C. & Bryn, K.Ø. 1965: Contributions to the Mineralogy of Norway. No. 31. Tysonite (fluocerite), a new mineral for Norway. Norsk Geologisk Tidsskrift 45 , 177-188.
Taylor, B.E. & Friedrichsen, H. 1983a: Oxygen and hydrogen isotope disequilibria in the Landsverk I pegmatite, Evje, southern Norway; evidence for anomalous hydrothermal fluids. Norsk Geologisk Tidsskrift 63, 199-209.
Taylor, B.E. & Friedrichsen, H. 1983b: Light stable isotope systematics of granitic pegmatites from North America and
Norway. Chemical Geology 41, 127-167.

N G F A bstracts and Proceedings, N o. 2, 2003
85
Taylor, S.R., Heier, K.S. & Sverdrup, T.L. 1960: Contributions to the Mineralogy of N orway, 5. Trace element variations in three generations of feldspars from the Landsverk I pegmatite, Evje, southern Norway. Norsk Geologisk Tidsskrift
40, 133-156.
Tibballs, J.E. & Olsen, A. 1977: An electron microscopic study of some twinning and exsolution textures in microcline amazonites. Physics and Chemistry of Minerals 1, 313-324.
Uebel, P.-J. 1977: Internal structure of pegmatites, its origin and nomenclature; derived from petrographic studies of steeply dipping dikes in South-West Africa and extended to gently dipping dikes observed in southern Norway. Neues Jahrbuch für Mineralogie Abhandlungen 131, 83-113.
Vogt, J.H.L. 1910: Norges jernmalmforekomster. Norges Geologiske Undersøkelse 51, 225 pp.
Vogt, J.H.L. 1918: Jernmalm og jernverk. Norges Geologiske Undersøkelse 85, 181 pp.
White, A.J.R. & Chappell, B.W. 1977: Ultrametamorphism and granitoid genesis. Tectonophysics 43, 7-22.
Wyllie, P.J. & Tuttle, O.G 1960: The system CaO – CO – H O and the origin of carbonatites. Journal of Petrology 1, 1-46.

86
N G F A bstracts and Proceedings, N o. 2, 2003