Journal of Conference Abstracts

Session O08:4A

O08 : 4A/01 : H2

Reconstructing the Liquid Line of Descent of the Skaergaard Intrusion

Michael Toplis (mtoplis@crpg.cnrs-nancy.fr)

CRPG-CNRS, Vandoeuvre-lès-Nancy, F-54501, France

The Skaergaard intrusion, East Greenland, has held an important place in the petrologic literature ever since its first description in 1939, often being cited as key field evidence for Fenner's proposal that basaltic differentiation may lead to liquids strongly enriched in iron with little or no enrichment in silica. However, for the last ten years controversy has surrounded the compositional evolution of the Skaergaard liquid since Hunter and Sparks (1987) proposed that the appearance of cumulus magnetite resulted in strong silica enrichment and iron depletion of the liquid, as observed in volcanic systems and experimental studies. On the other hand, a variety of evidence from the Skaergaard rocks including contemporaneous dyke compositions, iron content of plagioclase, and melting experiments (e.g. McBirney and Naslund, 1990; Brooks and Nielsen, 1990; Tegner, 1997) has been used to support the contention that the liquid from which the middle zone of the intrusion crystallized (after the appearance of cumulus magnetite), was rich in iron and low in silica, as originally proposed by Wager and coworkers (e.g. Wager, 1960). This lack of consensus is disturbing given that the compositional evolution of the liquid is clearly of fundamental importance for understanding the physical and chemical interactions between melts and crystals during fractionation. A novel method for estimating the compositional evolution of the Skaergaard liquid will be presented, which combines mass-balance constraints (average rock compositions) and major-element mineral melt equilibria, without making any assumptions concerning relative sub-zone masses, or the initial or final liquid composition. The results suggest that after magnetite saturation iron depletion of the liquid was minimal, but strong silica enrichment occurred. The calculated liquid line of descent is consistent with compositions of chilled margin rocks, contemporaneous dykes, as well as granophyric pods preserved in the upper zone of the intrusion. Furthermore, the liquid evolution is similar to independently modelled liquid lines of descent of ferrobasaltic systems under conditions closed to oxygen (Toplis and Carroll, 1996). This method may be applied to a range of other basaltic layered intrusions in order to estimate liquid lines of descent.

Brooks CK & Nielsen TFD, Contrib Mineral Petrol, 104, 244-247, (1990).

Hunter RH & Sparks RSJ, Contrib Mineral Petrol, 95, 451-461, (1987).

McBirney AR & Naslund HR, Contrib Mineral Petrol, 104, 235-240, (1990).

Tegner C, Contrib Mineral Petrol, 128, 45-51, (1997).

Toplis MJ & Carroll MR, J Petrol, 37, 837-858, (1996).

Wager LR, J Petrol, 1, 364-398, (1960).

O08 : 4A/02 : H2

Late Cretaceous Plutonic Rocks within the Ophiolitic Melange of Crete: Constraints on Their Origin and Evolution

Alexander Langosch

(langosha@gwp-min.min.uni-koeln.de)¹,

Heinz-Günter Stosch

(stosch@ipg003.bio-geo.uni-karlsruhe.de)²,

Eberhard Seidel

(E.Seidel@granat.min.uni-koeln.de)¹ &

Martin Okrusch

(okrusch@mail.uni-wuerzburg.de)³

¹ Institut für Mineralogie und Geochemie, Zülpicher Str. 49 b, 50674 Köln, Germany

² Institut für Petrographie und Geochemie, Kaiserstr. 12, 76131 Karlsruhe, Germany

³ Mineralogisches Institut, Am Hubland, 97074 Würzburg, Germany

The uppermost tectonic unit of the Cretan nappe pile consists of Jurassic ophiolites and an underlying ophiolitic melange. High-temperature crystalline slices form a prominent constituent of the melange. They are composed of upper amphibolite facies metamorphic rocks and metaluminous to slightly peraluminous plutonic stocks. All rock types yield Late Cretaceous K-Ar ages of approximately 70 Ma. Remnants of this formerly widespread HT belt crop out at several localities in Eastern and Central Crete and on some Aegean islands. Petrographic and geochemical features separate the plutonic rocks of Crete into two suites.

The magmatic rocks from Eastern Crete exposed at a few localities south of Agios Nikolaos range from gabbrodiorite to granite with a clear predominance of diorites and quartz diorites. The country rocks are represented by pure marbles and siliceous marbles. Sharp intrusive contacts to the country rocks and an excellent preservation of igneous textures are typical characteristics of this suite which defines a calc-alkaline trend on an AFM diagram. Element variations indicate that fractionation (of hornblende, plagioclase, ilmenite, apatite, biotite and finally K-feldspar) was the major process during magma evolution.

The main outcrop area of plutonites in Central Crete is located in the western part of Asterousia Mountains near the south coast. The intrusive rocks are mostly granites and granodiorites while mafic samples are subordinate. They are all more or less deformed and some appear even as augen gneisses. The contacts to country rocks (including metapelitic gneisses, amphibolites, marbles) are diffuse in some cases, especially in the vicinity to migmatic metapelites. Evidence of high to moderate temperature deformation suggests a syntectonic intrusion. Element variations display some differences to the rocks from Eastern Crete. The calc-alkaline trend shown on an AFM diagram is shifted to a higher Fe/Mg ratio for the whole suite. Intrusive rocks from Central Crete are richer in Ti, P, Zr, Nb, Y, Zn and REE for a given SiO₂ content. Fractionation of pyroxene, plagioclase and further phases played an important role in the early magma evolution. Additionally, geochemical characteristics require a crustal contribution, probably by assimilation of metasedimentary rocks with abundant zircon.

Recent Rb-Sr data of both, igneous and metamorphic rocks confirm the Late Cretaceous age and support contemporaneous intrusions and metamorphism. Nd and Sr isotopic compositions reveal similarities between the two Cretan plutonic suites. Mafic samples have compositions in the upper range of arc basalts or the lower range of MORB while intermediate to felsic rocks occupy an array between MORB and sediments, consistent with their formation by an AFC process. The combined geochemical evidence suggests an active plate margin as the most likely tectonic setting for generating melts of the Cretan granitoids.

O08 : 4A/03 : H2

Geochemical Variations in the Clasts and Matrix of Polymict Peridotites

Lisa Morfi (lmorfi@glg.ed.ac.uk)¹,

Ben Harte (Ben.Harte@glg.ed.ac.uk)¹,

Peter Hill (Peter.Hill@glg.ed.ac.uk)¹ &

John Gurney (john.gurney@minserv.co.za)²

¹ Grant Institute of Geology and Geophysics, University of Edinburgh, West Mains Road, Edinburgh, Scotland

² Department of Geological Sciences, University of Cape Town, Cape Town, South Africa

Polymict peridotites are an exceptional suite of mantle xenoliths recovered from kimberlites in Kimberley, South Africa. They have breccia-like characteristics and contain rock and mineral fragments derived from the mantle. These fragments are embedded in a very variable matrix often rich in phlogopite together with other silicates, ilmenite, rutile and sulphides. A two stage petrogenesis has been suggested for these xenoliths. This involves creation of the polymict rocks through sampling of the upper mantle by a fluid which migrated from depth and eventually crystallises the matrix assemblage. The consolidated polymicts rocks were later sampled and transported to a near-surface environment by a younger kimberlite intrusion. The polymict rocks provide an excellent opportunity to study aspects of melt infiltration and mantle metasomatism.

Major element analyses of orthopyroxene and phlogopite clasts in three of the xenoliths indicate core-rim variations. The orthopyroxene clasts have variable core compositions whereas the rims show a common Fe/Fe+Mg value of approximately 0.11. Often, individual orthopyroxenes show two rim compositions - an inner rim with higher Fe/Fe+Mg values and an outermost rim of slightly lower Fe/Fe+Mg values. This feature is not necessarily consistent within one nodule. Analyses of matrix orthopyroxenes in one of the nodules reveal that these crystals have a similar Fe/Fe+Mg value to the rims of the larger orthopyroxene clasts. The phlogopite clasts also show variable core compositions and two distinct rim compositions - both of higher Fe/Fe+Mg values than the cores. With respect to other elements, the phlogopite rims are consistently higher in Na, Cr, Al and Ti than the cores. This is not true of the orthopyroxene clasts. These features are suggestive of metasomatic processes operating on clasts derived from a variety of sources and preliminary interpretations involve the possibility of more than one pulse of a metasomatising fluid which has resulted in the heterogeneous rim compositions.

Trace element analyses of garnet clasts also show a variety of sources are present in the xenoliths. One polymict peridotite nodule contains small yellow-orange garnets intimately intergrown within a fine-grained matrix of orthopyroxene, phlogopite and ilmenite. The REE profile of these garnets is typical of megacrystic garnets from kimberlite sources. High TiO₂, low Cr₂O₃ and moderate CaO support this. A late-stage garnet overgrowth surrounding a large garnet fragment shows similar compositions to these matrix garnets. These features suggest that the melt which has transported the clasts and crystallised to form the matrix of the polymict peridotites is a relatively primitive ultrabasic melt of probably asthenospheric origin (Harte et al., 1993).

Harte B, Hunter RH, Kinny PD, Philosophical Transactions of the Royal Society of London A, 342, 1-21, (1993)

O08 : 4A/04 : H2

Subjotnian (1.53-1.50 Ga) Alkaline Magmatism in Central Sweden: Origin and Evolution of the Magmas

Ulf B. Andersson (ulf.andersson@geo.uu.se)

Institute of Earth Sciences, Uppsala University, Villavägen 16, SE-752 36 Uppsala, Suède

A number of minor Subjotnian intrusive complexes in central Sweden, belonging to the westernmost and youngest (1.53-1.50 Ga) part Fennoscandian rapakivi magmatism, contain alkaline rocks as a major constituent. The alkaline rocks comprise saturated and over-saturated syenites, some trachytic dykes, and silicic peralkaline granites and dykes. Gabbros and met- to peraluminous granites are also an essential part of these complexes. Trachyte dykes have been found, particularly associated with the small Nordsjö intrusion. They display trachytic flow structures and contain phenocrysts of anorthoclase, Fe-Ti oxides, augite, and apatite, providing evidence that syenitic magmas were originally present, precipitating these phases. The syenites are particularly high in K, Na, Ba, and Eu, compared with the associated rocks, suggesting feldspar accumulation. This is also evident from the REE patterns, which show positive Eu anomalies, in contrast to those from the trachytes. The relative Fe enrichment is high in the syenites (Fe/(Fe+Mg)=0.85-0.96, av. c. 0.92), compared with the met-peraluminous granites (<0.90), while the peralkaline silicic rocks are all higher than 0.95. The latter contain typically one feldspar, a hypersolvus mesoperthite with no Ca, and alkali-rich amphiboles (riebeckite and ferrorichterite). Field relations promted geochemical testing of the cogenetic relation between the syenites and peralkaline silicic rocks. Least-square major element modelling was perfomed, using different syenites or trachytes as parental magmas arriving at peralkaline residual compositions. The calculations suggest that this evolution can be achieved by 50-70% crystallization of 75-85% anorthoclase, 7-15% Fe-Ti oxides, 3-10% augite, 2-7% fayalite, and 1-2% apatite, depending on start and end compositions, consistent with the modal phenocryst mineralogy observed in the trachytes (except for fayalite which, on the other hand, is abundant in the syenites). Independent tests of the major element models can be obtained by trace element modelling using crystal/liquid partion coefficients. Such calculations yield overall very good agreement, particularly for the REE but also for the HFS elements. Poorer fits are noted for the feldspar-related elements Eu, Ba, and Sr, which is understandable considering the uncertainty in the degree of feldspar accumulation in the syenites.In conclusion: the phenocryst mineralogy observed in the trachyte dykes probably represents the crystallizing assemblage that drives magma evolution from saturated syenitic to peralkaline silicic composition. In contrast, no particular evidence suggests a cogenetic connection between the alkaline rocks and the associated basic or met-peraluminous granites. A probable scenario: three different protoliths (mantle, mafic lower crust, and felsic calcalkaline crustal) gave rise to three different primary magmas (gabbroic, syenitic, and metaluminous granitic) simultaneously in time, which cointruded along the same crustal channelways to their present emplacement levels.

O08 : 4A/05 : H2

F-Li-B-Rich Two Mica S-Monzogranite Cored at 3.5 km Depth in the Larderello Geothermal Field (Italy)

Giuseppe Cavarretta (gc@gea.geo.uniroma1.it)¹ &

Mariano Puxeddu²

¹ CNR - C.S. Quaternario E. A., Dip. Scienze della Terra, P.le A. Moro, 5, 00185 Roma, Italy

² CNR - Istituto Internazionale Ricerche Geotermiche, Piazza Solferino, 2, 56126 Pisa, Italy

The well Monteverdi 7 (MV7) crossed at 3.5 km depth a finegrained, heterogranular, hypidiomorphic, 3.8 Ma old, two mica S-monzogranite. Quartz, K-feldspar, plagioclase, biotite, muscovite, cordierite and and alusite are the main phases. Tourmaline, titanite, apatite, zircon and fluorite are the accessory minerals. A metapelitic xenolith is rich in sillimanite and green spinel. Two generations of plagioclase were recognized: early limpid intratelluric crystals (An₄₀-An₆₀) and late fine grained turbidones (An₄-An₂₃). K-feldspar occurs with perthitic hypersolvus albite-rich microphenocrysts and fine grained albite-poor crystals. Biotite is a siderophillite-rich variety plotting within the "clusters" field in the Mg vs. F diagram (Mason, 1992). Muscovite includes an early primary paragonite-rich magmatic type with FeO_tot=1-1.5%, MgO=0.8-1.1% and a paragonite-poor late magmatic one with FeO_tot=3-7%, MgO=1.5-4%; both types are F-rich (up to 0.95%). A gradual increase of fO₂ is observed from early primarymuscovites with Fe³⁺/(Fe²⁺ + Fe³⁺)=0.05-0.10 to late primary ones with Fe³⁺/(Fe²⁺ + Fe³⁺)=0.10-0.50. Fluorine intercept IV(F) for adiacent crystals of biotite and muscovite are 1.56 and 1.62 respectively, an evidence of equilibrium growth. Very high F contents characterize titanite (up to 4.66%) and apatite (up to 4.34%). Tourmalines plot in the field of Li-poor granitoid pegmatites and aplites (Cathelineau et al., 1992). Fluorite accur as anhedral blades (5-20 µm) associated with very late muscovite crystals (5-20 µm) within cavities, and along fractures and cleavages of calcicplagioclase.

The bulk composition of the MV7 rock is comparable to that of F-Li-B-rich peraluminous two-mica rhyolites and granites of Utah, Alaska, New Mexico, Peru, Australia, SW England, Portugal, Czech Republic, Mongolia and Himalaya. Liquidus temperatures of 850^o-825^oC and initial 4-5 wt% H₂O are hypothesized. The abundance of LiCl up to 30 wt% eq. in fluid inclusions (Cathelineau et al, 1994; Valori et al., 1992), the strong final enrichment in F and B testified by the occurrence of late tourmaline and fluorite, and the geothermometric indications supplied by the biotite-apatite pair and by the muscovite compositions suggest solidus temperatures as low as 540^o-600^oC and P_H2O=P_Load=115 MPa (calculated from rock densities in the MV7 well). The last residual melts were strongly enriched in Na as suggested by a local pervasive albitisation of the whole rock without any evidence of a hydrothermal post-magmatic nature of the albite crystals. Finally the downward percolation of fluorine-free meteoric waters led to chloritisation of biotite and sericitisation of feldspars.

Mason RA, Can Mineral, 30, 343-354, (1992).

Cathelineau M, Marignac C, Boiron MC, Gianelli G, Puxeddu M, Geochim Cosmochim Acta, 58, 1083-1099, (1994).

Valori A, Cathelineau M, Marignac C, J Volcanol Geotherm Res, 51, 115-131, (1992).

O08 : 4A/06 : H2

Multi-Generation, Multi-Source Dyke Swarms in the Area of the Central Bohemian Plutonic Complex

Franticek V. Holub (frholub@prfdec.natur.cuni.cz)

Inst. of Petrology and Struct. Geology, Fac. Science, Charles University, Praha 2, Czech Republic

The Central Bohemian Plutonic Complex (CBPC) of Lower Carboniferous age intruded into the Central Bohemian Suture Zone (CBSZ) separating the relatively cold Barrandian block at NW from the hot, rapidly exhumed metamorphic rocks of the Moldanubian block at SE. CBPC comprises various types of granitoid and less voluminous mafic rocks ranging in composition from calc-alkaline (CA) and high-K calc-alkaline to shoshonitic (HK) to ultrapotassic (UK).

Dyke swarms that are highly variable in composition, orientation and temporal relations to plutonic rocks intrude the area. Pre-CBPC dykes (N-S trending diabases, variously oriented spessartites, and felsic granitoid porphyries) are known from the adjacent parts of the Barrandian block only. Within major parts of CBPC itself, there occurs a conspicuous regional swarm of innumerable E-W to SE-NW trending dykes formed by multiple injections of highly diverse magmas. They correspond to dolerite, microdiorite and spessartite (CA), kersantite, granodiorite to granite porphyries (HK), minette (UK), melasyenite to melagranite porphyries (UK, probably of hybrid nature), etc. Many minettes and UK porphyries are well comparable in composition to and perhaps broadly coeval with the ultrapotassic plutonic rocks of the durbachite group (zircon ages about 343-346 Ma, Holub et al. 1997).

Among mafic dykes the minettes are the youngest and most widespread. Accompanied by UK porphyries, they occur even in a part of the Moldanubian adjacent to S margin of CBPC. Older small magma batches were probably unable to penetrate the hot, low-viscosity, anatectic Moldanubian crust. The "peri-Moldanubian" zone of CBPC inclusive of the durbachitic rocks and also the Moldanubian complex itself are intruded by numerous dykes of tourmaline leucogranites originated in a crustal source. They should be younger than almost all the "typical" dyke rocks, perhaps excepting some minettes and the youngest, N-S striking dyke swarm of highly evolved sub-volcanic granite porphyry (rhyolite).

Multiple intrusive pulses forming the dyke swarms should be derived from distinct mantle and crustal sources. The highly variable geochemistry of mafic dykes reflects a strongly heterogeneous nature of the lithospheric mantle, which was depleted to various degrees by previous extraction of basaltic melts and subsequently enriched in hygromagmatophile elements. High LILE/HFSE elemental ratios suggest enrichment by hydrous fluids released from a subducted slab due to its metamorphic dehydration. The highly refractive nature combined with the most pronounced enrichment are characteristic for the mantle sources of ultrapotassic magmas. Some systematic variations in geochemistry of comparable mafic dykes from both sides of CBSZ indicate its first-rate significance and a different compositional development of the two lithospheric mantle domains before the final juxtaposition of the Barrandian and Moldanubian blocks.

Holub FV, Rossi Ph, Cocherie A, CR Acad. Sci. Paris Sci. Terre. Planet, 325, 19-26, (1997).

O08 : 4A/07 : H2

Partially Molten Lower Crust beneath Northern Tibet Triggered by Underplating of Melts Derived from Lithospheric Mantle?

Sun-Lin Chung (sunlin@ccms.ntu.edu.tw)¹,

Ching-Hua Lo¹ &

Tung-Yi Lee (t44001@cc.ntnu.edu.tw)²

¹ Department of Geology, National Taiwan University, 245 Choushan Road, Taipei 106, Taiwan

² Department of Earth Sciences, National Taiwan Normal University, Taipei, Taiwan

The INDEPTH team has provided evidence for a partially molten layer (10-15 km) at the mid-crustal level in southern Tibet. Owens and Zandt (1997) reported seismic data suggesting that from south to north in the Tibetan plateau the crust thins by up to 20 km and pervasive partial melting of the lower half of the crust (>30 km) occurs in the north as a result of the India-Asia collision. An apparent drawback, however, is that in northern Tibet post-collisional, crustally-derived magmas, such as the leucogranites widespread in the Himalayas and southern Tibet, have been rarely identified. On the other hand, geochemical investigations of post-collisional potassic igneous rocks over the entire plateau have repeatedly suggested that partial melting has taken place in the underlying lithospheric mantle. The potassic magmatism may have resulted from convective removal (or "delamination" as often loosely referred) of the thickened Asian lithosphere, a process that could have also accounted for rapid uplift and development of north-striking normal faults in the Tibetan plateau.

We therefore envision that the proposed partially molten layer in the lower crust of northern Tibet does not necessarily only indicate melting of the lower crust. It may instead be interpreted in terms of extensive underplating of magmas derived from the lithospheric mantle, a process capable of upward influxing heat for the unusually high Poisson's ratios reported in the region. Whereas in the Himalayas and southern Tibet a thicker crust allows large scale crustal anatexis to form the leucogranites, delamination-induced partial melting in the north could have occurred preferentially in the lithospheric mantle because of a thinner crust (and lithosphere) and hotter geothermal structure. If substantial amounts of mantle-derived melts intrude into the lowermost part of crust, i.e., close to and above the Moho, they would be able to induce significant crustal melting. Then the abnormal lower crustal features observed in northern Tibet may be caused by combined effect of mantle and crustal melts. Considering the fact that so far crustally-derived igneous rocks have been identified only in a restricted area in central northern Tibet, a question remaining is why crustal melts do not erupt more widely, as did mantle-derived potassic rocks, or intrude into shallow depths and thus can be exposed by now. Perhaps crustal melts do exist in a thick range in the lower crust of northern Tibet, but they need more time to move upward and/or are awaiting a more effective mechanism to enhance melt migration and eruption.

O08 : 4A/09 : H2

The Kraemer Island Macrodyke, Skaergaard Region, East Greenland ­ Trapped Liquid Shift During Crystallization of a Tholeiitic Intrusion

Peter Momme (MOMME@geo.aau.dk) &

Richard Wilson (JRW@geo.aau.dk)

Geologisk Institut, Aarhus Universitet, DK-8000 Århus C, Denmark

The Kraemer Island Macrodyke (KIMD) is situated in central east Greenland, about 7 km west of the Skaergaard Intrusion. KIMD is exclusively exposed against Precambrian gneiss on Kraemer Island where it has a length of ~5 km and a width varying from 600 to 1000 m. The tholeiitic KIMD consists mainly of olivine gabbros, some with well developed clinopyroxene oikocrysts. KIMD is part of a dyke swarm that shows chilled margin whole-rock compositions similar to calculated Skaergaard liquids ('Skaergaard-like' dykes in Brooks & Nielsen, 1978). Weak modal layering (N-S/~30^o W) is developed in the northern outcrops on Kraemer Island. The most primitive clinopyroxene and plagioclase compositions in the gabbros (An₆₉ : Mg#cpx80) are slightly more primitive than in Skaergaard LZa. In contrast to clinopyroxene and plagioclase, olivine grains are unzoned and the most primitive composition is Fo₆₅. Clinopyroxene oikocryst cores are chadacryst-free and are believed to have grown from a large magma reservoir during the cumulus stage. Some KIMD olivine gabbros have abnormally high modal contents of olivine (20-25%) compared to the majority of the KIMD cumulates (5-10%). These olivine-rich gabbros occur near the eastern margin of the intrusion, suggesting that this could be a feature related to a relatively high degree of undercooling at the intrusion margin. In bivariate plots of major- and trace elements vs. wt% MgO, the compositional variance in the olivine-rich samples is covered by simple weighted average addition of olivine (Fo₇₈) alone to an appropriate magma composition (from the cogenetic 'Skaergaard-like' dyke swarm) inferring that accumulation of clinopyroxene and plagioclase did not play a role in formation of these rocks. This excludes the possibility that the olivine-rich gabbros simply are the result of crystallization of cotectic proportions of the mineral assemblage. Furthermore, Fo₇₈ coexists with the observed An₆₉ and Mg#₈₀ clinopyroxene in calculated "COMAGMAT" (Ariskin et al., 1993.) models of fractional crystallisation. In 'intra-plutonic chill zones' (ICZ) which occur near the western margin of KIMD new influxes of hot magma crystallized against pre-existing macrodyke walls to form olivine-rich gabbros. The olivine content here gradually decreases towards the intrusion centre to form 'normal' olivine-gabbros. Redistribution of olivines (initially produced at an ICZ) during a prolonged replenishment episode, could result in addition of olivine to a KIMD parental magma, making the observed compositional variations of the olivine-rich KIMD gabbros possible. All in all, combination of thin section observations, mineral- and whole-rock compositions as well as computer models suggest that olivine in KIMD has re-equilibrated to more Fe-rich compositions as a result of "trapped liquid shift" with Fo ~ 13 - 20 mol%.

Brooks CK & Nielsen TFD Lithos, 11, 1-14, (1978).

Ariskin AA, Frenkel MY, Barmina GS & Nielsen RL, Computers and Geosciences, 19, 1155-1170, (1993).

O08 : 4A/10 : H2

Separation of Bulk and Interstitial Liquid Fractionation Recorded by Pigeonite-Augite Intergrowths in the Graveyard Point Sill, Oregon

Gregor Markl (markl@ruf.uni-freiburg.de)¹ &

Craig White²

¹ Institut für Mineralogie, Albertstrasse 23 B, 79104-Freiburg, Germany

² Dept. of Geosciences, Boise State University, Boise, ID 83725, USA

The 150 m thick late Miocene Graveyard Point sill (GPS) is situated at the Idaho-Oregon border near the southwestern edge of the western Snake River Plain. It records from bottom to top continuous fractional crystallization of a tholeiitic parent magma (lower chilled border, FeO/(FeO+MgO)=0.59, Ni=90 ppm) towards granophyrs (late pods and dikes, FeO/(FeO+MgO)=0.98, 78 wt.% SiO₂ 3.5 wt.% K₂O, <4 ppm Ni) showing a typical trend of Fe and P enrichment. Fractionating minerals are olivine (Fo₇₉-Fo₂), augite (X_Fe=0.18-0.95), feldspars (An₈₀Or₁-An₁Or₆₂), Fe-Ti oxides (Ti-rich magnetite and ilmenite), apatite and in two samples pigeonite (Wo_18-28Fs_41-54). The granophyrs may bear some quartz. Compositionally zoned minerals record a large interval of the fractionation process in every single sample, but this interval changes with stratigraphic height. In pigeonite-bearing samples, olivine is scarce or lacking and because pigeonite occurs as characteristic overgrowths on augite, its formation is interpreted to be related to the schematic reactionaugite + olivine = pigeonite that defines the cotectic between augite and pigeonite in olivine-saturated basaltic systems. Line measurements with the electron microprobe reveal that the transition from augite to pigeonite is continuous. However, some crystals show an abrupt "reversal" towards augite after pigeonite growth. These observations appear to be similar to some textures described from lunar basalts, but they have not yet been described from terrestrial rocks. We interpret the textures to record the competition of two processes during fractionation in the GPS: fractional crystallization of the bulk liquid (the bulk melt separates from solids and interstitial liquids in the solidification front) and fractional crystallization of interstitial melt in the solidification front itself. Interplay between those two processes is proposed to account for the observed variations in mineral chemistry and mineral textures and is consistent with pigeonite stability field variations in published phase diagrams of basaltic systems during fractionation.

O08 : 4A/11 : H2

Petrography and Petrochemistry of the Upper Cretaceous Volcanic and Subvolcanic Rocks in Olucak Region (Gumushane-Torul/NE Turkey) and Their Genetic Relation to Torul Granitoide

Hakan Coban (coban@sdu.edu.tr)

Suleyman Demirel University Geology Dept. Isparta, Turkey

The study area includes the Olucak (Gumushane-Torul/NE Turkey) and its surrounding areas located at between the northern and southern zones of the East Pontide. Upper Cretaceous outcrops wide spreadly in the region and consist of magmatic rocks (Karaburun dacite-Torul granitoide) and volcano-sedimentary rocks (Kermutdere-Tepekoy formations). Two phases were determined in the Upper Cretaceous volcanism. The first phase (Early phase) includes the volcanics Kermutdere Formation. In this phase, rocks have basaltic-trachyandesite/tranhyandesite and phonolitic composition and indicate silica-saturated and silica-undersaturated magmas. The bottom level of the phase consist of the rocks having plagioclase (neutral) + sanidine + hornblende + clinopyroxene (augite). The upper part of the phase is composed of the rocks having plagioclase (acid-neutral) + sanidine + nepheline + sodalite (nosean) + clinoamphibole (arfvedsonite) and olivine. The second phase (late stage) is chracterized by a silica over-saturated magma and indicates rhyolitic to dacitic rocks in composition (Tepekoy formation). The rhyolites in the second phase have a perlitic texture and are a rapidly cooled magma origin. Some textural features, observing in Upper Cretaceous volcanites, such as melting inclusions in plagioclases, occurrence of needle-shaped apatites, ocellar texture (maphic mineral inclusion bearing quartz) are the evidences of the same aged homogene magma-mixing between both maphic and felsic magmas. The petrochemical analysis of volcanics showed that the volcanics of Kermutdere have alkaline (shoshonitic)-calc-alkaline magma character and the Karaburun dacite-Tepekoy formation volcanics have calc-alkaline magma character. These characters are associated with their mineralogic composition. These petrographic-petrochemical aspects and especially petrogenetic trace elements parameters of Upper Cretaceous volcanics indicate that volcanics are derived from different magma origines and Kermutdere formation volcanics have hybridic magma genesis related to partial melting of upper mantle and partially crustal material. Tepekoy formation and Karaburun dacite have dominantly crustal genetic magma and are interpreted that as Karaburun dacite and Tepekoy formation volcanics are the marginal sub-volcanic and volcanic extrusion facies of the Torul granitoide, which intruded in the Upper Cretaceous age.

O08 : 4A/12 : H2

Mantle Fluids and Genesis of Explosion and Impact Circulars Structures. Problems of Impact Volcanism

Vladimir Sakhno (olegavchenko@geocities.com)

Kirova 70-126, Vladivostok, Russia

Analyses of geological, petrologic and geochemical data on the El'gygytgyn caldera in the Eastern Chukotka and the Popigai and other circular structures in the Siberian North of Russia, which are assumed as astroblemes, revealed that they all were of endogenic origin. There is regularity in their development - all are located in the center of large arched uplifts composed of volcanic rocks. Processes of the structure formation was associated with volcanism of acid-to-basic character, took long interval from Early Cretaceous to Cenozoic and culminated in a powerful explosion in the center of uplift and caldera formation. Endogenic impactite and impact breccia, slag and glass characterize this sage. Rocks resulted during the explosion usually have mineral associations with planar features completely or partly replaced by diaplectic glass. The bulk of the impact melt is represented by fresh glass with fluidal texture separate sprites of which are differ from each other by color and composition. Planar features and diaplectic glasses in minerals as well as finding of coesite permitted defenders of the astrobleme theory to consider these structures to be impact, produced by impact of cosmic bodies. Thorough investigation of the rocks has not revealed any geochemical anomalies, whilst study of liquid phase of minerals and rocks revealed a high content of H and a high hydrogen-carbon ratio. The relation was close to 2. Rocks before the caldera stage (pre-impact explosion) have the marked predominance of carbonic oxides in fluid. The simulation of chemical reactions of gases with the "Selector" program (Karpov et. al., 1997, 1998) has demonstrated that the temperature and pressure these rocks having been formed under were more than 2000°C and 200 kb respectively. That did not correspond to the real adiabatic conditions of acid melts generating in the focuses of volcanism, but conform to conditions of gaseous chemical reactions developing through the mechanism of chain reaction of explosive mixture (hydrogen with oxygen) with the spasmodic velocity increase towards the higher values. That brings to a powerful explosive release of energy, high temperature and pressure, to an occurrence of minerals high in density (diamond, coesite) and minerals with planar features. Isotopic geochemistry of fluid as well as petrologic characteristics of the rocks supports the possibility of impact origin of the circular structures. It is also probable that impact volcanism was of widespread occurrence throughout the Earth's history (Marakushev et. al., 1993, 1997, and others).

O08 : 4A/13 : H2

Oxygen Isotope Composition of Plume-Related Magmatism in the Manus Basin, S.W. Pacific

Colin G. Macpherson (c.macpherson@gl.rhbnc.ac.uk)¹,

David R. Hilton (drhilton@ucsd.edu)²,

David P. Mattey (d.mattey@gl.rhbnc.ac.uk)¹ &

John M. Sinton (sinton@soest.hawaii.edu)³

¹ Department of Geology, Royal Holloway University of London, United Kingdom

² Geosciences Research Division, Scripps Inst. Oceanography, University of California San Diego, USA

³ Department of Geology and Geophysics, University of Hawaii, USA

The Manus Basin is a site of extension in the eastern Bismarck Sea, north of the active New Britain Arc. Magmatism occurs along the north-east striking Manus Spreading Centre and east striking Extensional Transform Zone. Smaller rift segments to the south and east are also volcanic. Like many other back-arc basins recent magmatism in the Manus Basin displays a range of petrogenetic affinities. Basaltic glasses form a spectrum between compositions typical of mid-ocean ridges and rocks that are arc-like. However, helium isotope analyses have revealed the presence of an additional component derived from a deep mantle plume.In this study we present a comprehensive oxygen isotope study of back-arc magmatism in the Manus Basin. ¹⁸O/¹⁶O ratios of basic glasses display a negative correlation with helium isotopes ratios. Higher ¹⁸O values are associated with ³He/⁴He ratios typical of the upper mantle. These high ¹⁸O values result from interaction between residual upper mantle and fluids recycled through the subduction zone. Glasses with high ³He/⁴He ratios, characteristic of a mantle plume, have ¹⁸O values that are lower than average mid-ocean ridge basalt ¹⁸O/¹⁶O ratios. Crustal contamination can not produce the observed correlation suggesting an ¹⁸O-depleted reservoir is associated with the mantle plume. This reservoir may represent parts of recycled oceanic crust that were hydrothermally altered at high temperatures. Alternatively, interactions between silicate and iron at the core mantle boundary may generate an ¹⁸O-depleted reservoir. Entrainment of such material in a plume, upwelling from the base of the lower mantle, could explain the relationship between the He and O isotope systems in the Manus Basin.

O08 : 4A/14 : H2

Zoning Patterns in Volcanic Feldspars as a Clue to Magmatic Processes: A New Approach Using Electron Microprobe

Catherine Ginibre (cginibr@ugcvax.dnet.gwdg.de) &

Gerhard Wörner (gwoerne@gwdg.de)

Geochemisches Institut, Goldschmidtstr. 1, D-37077 Göttingen, Germany

Zoning in phenocrysts, especially plagioclase, has been used as a tool to understand the history of the magma chamber in which they grew. However, the relationships between observed zoning patterns and processes and parameters which control them (growth, resorption, change in temperature, pressure or composition in the magma) are not well established. Here we investigate feldspars phenocrysts from Parinacota (Chile) and Laacher See (Eifel, Germany) volcanoes in order to understand these relationships. We use a combination of textural observation (optical microscopy and BSE images) and electron microprobe analysis of major, minor and trace elements and element mapping (Ca, Na, K, Al, Si, Ba, Sr, Fe, Mg, Ti).

A large number of feldspar crystals were investigated and show great diversity in textures (oscillatory zoning, resorption surfaces, and patchy zoning) but also some common features. Oscillatory zoning is present in most crystals and were probably destroyed by later dissolution and diffusion in others. Grey-value profiles across BSE images of oscillatory zoned crystals are representative of An content and show high spatial resolution (sub-µm). Zoning parameters which vary along a profile (or between crystals) are wave length, amplitude and shape of the oscillations as well as shape of growth zones boundaries (former crystal surfaces?). An aim of the study is to relate these oscillation parameters to growth conditions or igneous events in the magma chamber. Different styles of resorption surfaces correlate with different styles of zoning: low amplitude high frequency resorption surfaces are typical for oscillatory zoned plagioclase while high amplitude low frequency surfaces correlate with patchy zoning. Zoning and resorption patterns in sanidine are similar to those in plagioclase but less pronounced and larger scaled.

Minor and trace elements do not neccesarily correlate with each other and with major elements. For example, Ba and An-content correlate while Sr and An do not, suggesting fast diffusion for Sr in the crystal. Mg and Fe increase strongly but An decreases at the outermost rims which probably grew fast after eruption. This may indicate kinetic effects. Diffusion and growth kinetics are time dependent and thus may provide information for rates of processes and histories of the crystals in the magma chambers.

Session O08:4B

O08 : 4B/25 : H2

Eclogite and Granulite Nodules from the Mbuji Mayi Kimberlites (Kasai, R.D. Congo): Products of the Isobaric Cooling of an Underplated Mafic Gabbronoritic Complex

Daniel Demaiffe (ddemaif@ulb.ac.be)¹,

Saïd El Fadili (selfadil@ulb.ac.be)¹,

Luc Andre (lucandre@africamuseum.be)² &

Nicolas Coussaert

¹ Universitè Libre de Bruxelles, Géologie CP160/02, 50, Avenue Roosevelt B 1050 Bruxelles, Belgium

² Musée Royal de l'Afrique Centrale, Département de Géologie, 3080 Tervuren, Belgium

The Mbuji Mayi diamondiferous kimberlites have intruded (71 Ma ago) the ArcheanKasai craton that consists of high-grade migmatitic gneisses and a huge metamorphosed gabbronoritic complex.

Four main groups of rocks have been distinguished petrographically among the 160 deep-seated nodules : 1) bimineralic eclogites: omphacite + garnet ± accessory minerals; 2) kyanite-bearing eclogites: kyanite occurs either as symplectitic intergrowths with omphacite or as subhedral laths included in garnet; 3) diamond-bearing eclogite (one sample); 4) mafic granulites characterized by primary granulite assemblage (plagioclase +diopside +garnet) and secondary eclogite assemblage which consists of garnet and kyanite-omphacite intergrowths similar to those observed in kyanite eclogites. The nodules were variously affected by metasomatic process(es).

The bimineralic and kyanite-bearing eclogites were equilibrated under similar P-T conditions: 861-1091°C; 17-30 Kb and 815-946°C; 20-25 Kb respectively. The kyanite-bearing granulites were equilibrated under the same conditions (920°C and 20-21 Kb).

Detailed petrographic observations and mineral chemistry of the granulites show that the kyanite-omphacite intergrowths result from the breakdown of plagioclase (An₄₈) reacting with Al-rich diopside. This reaction possibly proceeded by isobaric cooling at relatively high P. The coexistence of relict diopside and newly formed omphacite and of two chemically distinct garnets in the granulites together with partial breakdown ofplagioclase suggest that these rocks have recorded the eclogite-granulite facies transition under the Kasai craton.

Major element compositions of bimineralic eclogites are broadly basaltic (Mg#= 42-72). Kyanite-bearing eclogites and kyanite-bearing granulites are peraluminous (17-28 and 21-24.4 wt% Al₂O₃) which suggests that their protoliths were plagioclase-rich (cumulate gabbros and/or anorthosites) This is confirmed by positive Ba, Sr, Pb and Eu anomalies in the trace element diagrams; these anomalies are higher for kyanite-bearing eclogites and kyanite-bearing granulites than for bimineralic eclogites.

The initial (at 2.5 Ga) Sr and Nd isotopic composition of the least metasomatized eclogites and of the kyanite-bearing granulites are similar (0.70102-0.70175; <epsilon>Nd= +2.80-+5.97) and comparable to those of the gabbronoritic complex (0.70156-0.70186; <epsilon>Nd= +1.3-+5.40) that outcrops 100 km to the SW of the Mbuji Mayi area. The isotopic data indicate a slightly depleted mantle source-region. The estimated age of the eclogite metamorphism (2.65±0.26 Ga; Sm-Nd on 5 ga-cpx pairs) is, within error limits, similar to the age of granulite metamorphism (2.4±0.1 Ga) that affected the gabbronoritic complex of Kasaï.

The Mbuji Mayi granulite and eclogite nodules could be the deep lower crust and the uppermost mantle equivalents of the metamorphosed gabbronoritic and anorthositic complex. Prolonged cooling, at high P, of this complex could generate the granulites and the two main groups of eclogites : the bimineralic eclogites being derived from protoliths of gabbronoritic compositions and the kyanite-bearing eclogites from gabbroic anorthosites and/or anorthosites.

O08 : 4B/26 : H2

Composition and Processes of the Mantle Lithosphere in Northeastern Brazil and Fernando de Noronha: Evidences from Mantle Xenoliths

Giorgio Rivalent (riva@unimo.it)¹,

Maurizio Mazzucchelli (mazzuc@unimo.it)¹,

Vicente A.V. Girardi (girardi@usp.br)²,

Maria A. Barbieri¹,

Alberto Zanetti (zanetti@crystal.unipv.it)³,

Steve L. Goldstein (steveg@ldeo.columbia.edu)⁴,

¹ Dipartimento di Scienze della Terra, Università di Modena, Italy

² Instituto de Geociencias, Universidade de Sao Paulo, Brazil

³ CNR, Centro di Studio per la Cristallochimica e la Cristallografia, Pavia, Italy

⁴ Lamont-Doherty Earth Observatory, Columbia University, Palisades, USA

Cenozoic alkali basalts in the Rio Grande do Norte State and the Fernando de Noronha Island contain abundant spinel facies mantle xenoliths. At Pico Cabuji (NE Brazil) both protogranular and porphyroclastic xenoliths occurr. The former mainly consists of lherzolites and the latter of harzburgites, with equilibrium temperatures of 1064±6°C and 1238±7°C, respectively. Pressure is assumed essentially equal in the two groups (~1.8 GPa). Porphyroclastic xenoliths have more refractory bulk rock and mineral phase compositions than the progranular ones, in keeping with their mode. Both groups show chemical variation trends consistent with fractional melting and basalt extraction. REE profiles of clinopyroxenes vary from LREE enriched (spoon shaped) to LREE depleted in the progranular group, whereas they are dominantly slightly convex upwards in the porphyroclastic clinopyroxenes. HFSE (Ti and Zr) anomalies are in general modest in the clinopyroxenes of both groups. At Fernando de Noronha, xenoliths are variably-textured lherzolites and harburgites. Their chemical variation trends overlap and extend those of the Pico Cabuji samples. Average temperature is 1056±87 and pressure 1.6±0.2 GPa. Clinopyroxenes have spoon-shaped to LREE depleted profiles similar to those of the Pico Cabuji protogranular type, but have higher REE concentrations and more marked negative HFSE spikes. Nd and Sr isotopes of the Pico Cabuji porphyroclastic clinopyroxenes (¹⁴³Nd/¹⁴⁴Nd = 0.51339 - 0.51255, ⁸⁷Sr/⁸⁶Sr = 0.70275 - 0.70319) and of Fernando de Noronha (¹⁴³Nd/¹⁴⁴Nd = 0.51323 - 0.51285, ⁸⁷Sr/⁸⁶Sr = 0.70323 - 0.70465) plot on distinct arrays originating from a similar, isotopically depleted composition and trending to lowNd-lowSr and lowNd-highSr, respectively. ¹⁴³Nd/¹⁴⁴Nd decreases, and ⁸⁷Sr/⁸⁶Sr increases, with increasing mg# of clinopyroxene and with all the parameters indicative of previous melting episodes (modal clinopyroxene, bulk rock MgO) The opposite correlations are observed with LREE concentration, La_n/Sm_n, and with all the parameters reflecting metasomatic enrichment. As a whole, chemical and isotopic data indicate that two isotopically and chemically distinct agents acted at Pico Cabuji and Fernando de Noronha and that the second is consistent with the presence of recycled crust in its source. Protogranular xenoliths from these two occurrences underwent chromatographic enrichment by porous flow percolation at low melt/rock ratio. At Pico Cabuji, porphyroclastic xenoliths are likely to be produced by fluid-assisted melting in shear zones of a former protogranular mantle. It is concluded that metasomatic processes have affected a heterogeneous lithospheric mantle with a depleted MORB-like composition. The metasomatic end-members identified in the xenoliths are recognized also in the host basalts.

O08 : 4B/27 : H2

CO₂ Rich High Pressure Melts Included in Minerals from Olivine Pyroxenite Xenoliths (West Eifel, Germany): Experimental and Ion Probe Study

Valentina Batanova (batanova@gs.ucsd.edu)¹,

Hans Seck (H.Seck@min.uni-koeln.de)²,

Alexander Sobolev (asobolev@glas.apc.org)¹,

Markus Klein (kleinm@geocip.geo.uni-koeln.de)² &

Marc Chaussidon (chocho@crpg.cnrs-nancy.fr)³

¹ Vernadsky Institute of Geochemistry, Russian Academy of Science, Kosigin 19, Moscow, 117975, Russia

² Universitaet zu Koeln, Institut fuer Mineralogie und Geochemie, Zuelpicher Str. 49b, Koeln, D-50674, Germany

³ CRPG-CNRS, BP 20, Vandoeuvre-les-Nancy Cedex, 54501, France

Abundant magmatic inclusions have been found in mineral phases of cumulative olivine pyroxenite xenolith (DW-905) occurring in the ash-tuffs of the Dreiser Weiher maar-type volcano, West Eifel, Germany (Witt-Eickschen & Kramm, 1998). Here we report the results of complex study of these inclusions by high temperature experiments both at low (1 atm, optical heating stage) and high pressures (10-12 kb piston cylinder apparatus). Quenched melt inclusions were analyzed by electron probe (major elements) and by ion probe (trace elements, H₂O and C) techniques.

Silicate melts, high density CO₂, sulfide melt and their combinations occur as inclusions in well-defined planes along healed intragranular fractures in olivine and clinopyroxene. These inclusions differ significantly in composition from secondary inclusions, which correspond to transporting melanephelinite. Thus they are interpreted as pseudosecondary inclusions trapped in xenolith minerals before its transportation. Abundant Cpx inclusions in olivine and olivine inclusions in Cpx are also present.

T-P conditions of formation of xenolith minerals estimated from the compositions of Cpx inclusions in olivine are as following: T=1200-1250°C; P more than 10-15 kb. These correspond to the temperature (T=1190-1250°C) estimations from the partial and full homogenization of melt inclusions and minimum pressure obtained from the density of CO₂ fluid inclusions (more than 9 kbar).

Compositions of melt inclusions corrected for quench olivine crystallization on the walls of cavity correspond to alkali rich highly silica undersaturated melts. They show significant range in respect to silica contents (SiO₂=33-41 wt.%), concentration of alkalis, Ca and P covering and exceed (toward lower Si and higher Ca, P and alkalis) all entire compositional range of reported Eifel magmas (Mertes & Schmincke, 1985). High sodium (up to 11 wt.% Na₂O) compositions not reported yet are also present. All inclusions show extreme enrichment (up to 1500X chondrite for Ba) by highly incompatible elements (light REE, Ba, Nb and Sr) and are relatively depleted in HREE (10-25 X chondrite for Yb). Again, as for major elements, trace elements concentrations cover all entire range reported for Eifel magmas (Mertes & Schmincke, 1985) and exceed it toward higher concentrations of all highly incompatible elements. Preliminary ion probe study of quench melts shows that major volatile element in the inclusions is C with concentration level exceeds 10000 ppm. H₂O concentrations in melts are lower than 0.50 wt%. Chlorine varied in the range of 0.2-0.4 wt%, sulfur in the range of 0.10 to 0.30 wt%.

Obtained data suggests that studied inclusions correspond to the melts of variable composition coexisting with olivine and clinopyroxene, CO₂-rich fluid and sulfide melt at temperature 1200-1250°C and pressure more than 10-15 kbar. Low viscosity of these melts and their extreme enrichment in highly incompatible elements make them effective metasomatic agents in lithospheric mantle.

Witteickschen G & Kramm U, Contributions to Mineralogy & Petrology, 131, 258-272, (1998).

Mertes H & Schmincke HU, Contributions to Mineralogy & Petrology, 89, 330-345, (1985).

O08 : 4B/28 : H2

Experimental Petrology of the Kiglapait Intrusion, Labrador, at 5 Kilobars

S. A. Morse (tm@geo.umass.edu)¹,

B. A. Sporleder (bsporled@u.washington.edu)²,

J. B. Brady (jbrady@science.smith.edu)³ &

A. L. Peterson (peterson@geo.umass.edu)¹

¹ Geosciences, UMass, Amherst, MA, USA

² Geological Sciences, Univ. Washington, Seattle, WA, USA

³ Geology, Smith College, Northampton, MA, USA

The liquid line of descent (LLD) has been determined for most of the Kiglapait intrusion. Samples were made up from analyzed whole rocks and separated minerals, adjusted to yield the observed crystal compositions at a given stratigraphic level, e.g. An₆₇, Fo₇₄ at the start of crystallization. Samples were melted and crystallized in graphite in piston-cylinder apparatus with pressure monitored by the solidus of sanidine. Experimental liquidus temperatures range from 1245°C to 1180°C in the Lower Zone (LZ), and from there down to a final solidus at 1065°C at the top of the Upper Zone (UZ). The LLD was calculated from cotectic glasses as oxygen norms. For the LZ it runs parallel to the modal track (Morse, 1979) from OL:AUG 80:20 at the start to 38:62 at the end, where saturation with AUG occurs. The path is slightly concave toward PL, as in the modal case. The provisional UZ path shows a steady increase in FSP content. The long LZ path over a small T range shows that large volumes of troctolite are produced by such a magma; both the lever rule and the volume estimates agree that augite saturation occurs at about 80% crystallized. Calculation with the MELTS subroutine gives only pyroxenes + PL, without OL, unless run at low pressure. The contrast with the experimental results indicates deficiencies in the model. Partition coefficients for Sr are 1.63 to 1.31 (An₆₈ to An₅₆), in strong contrast to a 1-atm study (Blundy, 1997) in which the SIMS analyses fail of mass balance in melts of known composition. For K, the D's are 0.39, 0.85 over the same PL composition range, much higher than any known 1-atm determinations. D for FeO in PL is 0.04, as commonly found. Ca was transferred rapidly into OL from PL in sintered OL+PL runs, showing the facility with which OL can gain or lose Ca against PL alone. The new data form the foundations for forward modelling of the intrusion history to compare with observed rocks. So far, most older models look quite good.

Morse SA, J. Petrol, 20, 591-624, (1979).

Blundy JD, Ch. Geol, 141, 73-92, (1997).

O08 : 4B/29 : H2

Experimental Investigation of Titanite Solid Solution Ca(Ti,Al)(O,F)SiO₄ in the Assemblage Titanite-Anorthite-Fluorite

Ulrike Troitzsch (ulrike@geology.anu.edu.au) &

David Ellis

Australian National University, Geology Department, Canberra A.C.T. 0200, Australia

The exchange reaction Ti⁴⁺ + O^2- = Al³⁺ + F^- in titanite has long been recognised to be strongly dependent on pressure, temperature as well as fluorine fugacity. Even though the presence of F has the potential to significantly increase the stability of titanite, especially towards higher pressure and lower temperature, most theoretical and experimental studies of phase equilibria have focused on pure Al- and F-free titanite, thus limiting their application to F-free environments. The only previous experimental study of AlF-rich titanite we are aware of (Smith, 1981) was carried out in an unbuffered, supersolidus assemblage, thus not allowing for the separation of the influence of fluorine fugacity on the titanite composition from that of pressure and temperature. We are investigating the pressure and temperature dependence of the Al-content of titanite in the subsolidus assemblages titanite-fluorite-anorthite (between 5 and 20 kbar) and titanite-fluorite-zoisite (between 20 and 35 kbar). The entire range of binary titanite solid solution [Ca(Ti,Al)(O,F)SiO₄] has been synthesised. Titanite coexisting with anorthite and fluorite reaches Al-contents up to X_Al=0.6 [X_Al=Al/(Al+Ti)], that coexisting with zoisite up to X_Al=1, thus clearly exceeding the maximum Al-content X_Al=0.53 of the previous experimental study (Smith, 1981). Since we were able to determine the change in unit cell volume of titanite with X_Al using X-ray diffraction data, the reaction anorthite + fluorite = CaAlFSiO₄ has the potential to yield thermodynamic data for CaAlFSiO₄, the Al-F end-member of titanite solid solution, as well as activity-composition relationships for intermediate compositions. This can then be applied to other equilibria involving titanite, and represents a first step towards the application of phase relations of titanite determined in experimental, F-free environments to F-bearing rocks. Phase equilibria including zoisite at higher pressure are not suited for the extraction of thermodynamic data for CaAlFSiO₄, because the zoisite in these experiments contains considerable amounts of F (about X_F=0.5 [X_F=F/(F+OH)); however, thermodynamic data for fluoro-zoisite are not available.

Smith DC, Progress in Experimental Petrology N.E.R.C. Publication Series, D-18, 193-197, (1981).

O08 : 4B/30 : H2

Low-Pressure Larnite-Normative Trend of Melt Evolution: Evidence from Synthetic Systems and Melt Inclusions

Leonid D. Krigman (elkor@geokhi.msk.su)¹,

Robert A. Ishbulatov²,

Troels F. D. Nielsen (nielsent@dlc.ku.dk)³ &

Ilya V. Veksler (ivv@dlc.ku.dk)³

¹ Vernadsky Institute of Geochemistry, Kosygin str. 19, 117975, Moscow, Russia

² Institute of Experimental Mineralogy, Chernogolovka, Noginsk region, Russia

³ Danish Lithosphere Centre, Oester Voldgade, 10L, Copenhagen, 1350, Denmark

Normative larnite in an igneous rock requires a silica-undersaturated bulk composition with high CaO/Al₂O₃ and CaO/MgO. In terms of mineral composition the rocks are usually melilite-bearing and, in some cases, approach monomineralic melilitolites. Larnite-normative liquids are documented in chilled dykes and lavas and by data on melt inclusions.

The vast majority of natural basaltic liquids do not evolve to larnite-normative compositions, because low-pressure crystallization of common SiO₂ - saturated Ca-silicates (e.g., plagioclase and diopsidic clinopyroxene) decreases the CaO content of the liquid and drives it to SiO₂-rich, granitic residuals. Thus, a SiO₂-undersaturated (e.g., melanephelinitic) parental liquid in which plagioclase is unstable is the first requirement for the larnite-normative trend.

The simple and basic model which demonstrates how low SiO₂ activity in a dry system results in the larnite-normative trend is the well-studied Nepheline (NaAlSiO₄) - Diopside (CaMgSi₂O₆) join. Reaction between the components drives the liquid to larnite-normative compositions, but a further increase in CaO in the liquid is soon halted by the onset of melilite crystallization.

The larnite-normative trend is greatly enhanced in the presence of water and fluorine. Stabilization of phlogopite and/or amphibole in equilibrium with SiO₂-undersaturated liquids provides an effective mechanism to produce highly larnite-normative, melilitic residual liquids. We have demonstrated this in experimental studies of several joins in the SiO₂-undersaturated part of the F-doped kalsilite-based normative tetrahedron (Veksler et al., 1998). Because F-phlogopite is much more stable than the natural OH-analogue, the F-doped system presents an extreme case of maximum phlogopite stability and its maximum effect on the melt evolution. We examined the effects of F-OH substitution by the study of the Diopside - Phlogopite (KMg₃AlSi₃O₁₀(OH)₂) and Diopside - Kalsilite (KAlSiO₄) - H₂O joins at 0.2 GPa. The preliminary results show that despite significant topological differences with the analogue F-doped join, the larnite-normative trend is observed in the liquids of the Diopside - Kalsilite - H₂O join.

CO₂/H₂O values are another important characteristic of natural melts which may have important consequences for the larnite-normative trend. Melilitolites are often found in close association with carbonatites and both rock types demonstrate extreme enrichment in CaO. Data on melt inclusions (Nielsen et al., 1997) suggest that carbonatitic melts may be derived from carbonated melilitic liquids by liquid immiscibility. It seems that the nepheline+diopside (ijolitic) association is an important branching point in the magmatic evolution and a subtle balance of total pressure, CO₂/H₂O and other conditions determine the evolution towards melilitolite and minor immiscible carbonatite, or to residual carbonatitic liquids associated with melilite-free ijolitic associations.

Nielsen TFD, Solovova IP & Veksler IV, Contrib. Mineral. Petrol., 126, 331-344, (1997).

Veksler IV, Fedorchuk YM & Nielsen TFD, Contrib. Mineral. Petrol, 131, 347-363, (1998).

O08 : 4B/33 : H2

The Use of Iron Partitioning between Plagioclase and Melt as an Oxygen Barometer for Hydrous Magmatic Systems

Max Wilke

(max.wilke@mineralogie.uni-hannover.de) &

Harald Behrens

Institut für Mineralogie, Universität Hannover, Hannover, Germany

Oxygen fugacity (f(O₂)) may have strong influence on the petrogenesis of magmatic rocks. Redox conditions in natural magmatic systems are mostly bracketed with the help of coexisting Fe-Ti oxides found in the rocks and experimental determination of phase relations (e.g. Rutherford & Devine 1996, Martel et al., 1998). The aim of this study is to add a further possible constraint on the f(O₂) by the use the f(O₂) dependant partitioning behaviour of Fe between plagioclase and melt. Crystallization experiments were performed using synthetic tonalitic glasses at 500 MPa water pressure and 750/850°C with varying f(O₂). The obtained partition coefficients (D_Fe) for iron vary from 0.54±0.12 for oxidizing conditions (Cu-Cu₂O) to 0.085±0.013 for reducing conditions (WM - 0.1). At 750°C the partition coefficients increase at constant f(O₂). A model is developed to describe the dependance of the partition coefficients on f(O₂), which is based on the redox equilibrium of ferric and ferrous melt species and describes the apparent partition coefficient as a linear combination of the two hypothetical endmember-partition coefficients of the ferric and ferrous cation. By fitting the experimental data to the model the parameters D(Fe³⁺), D(Fe²⁺) and K_m can be determined, where K_m denotes the equilibrium constant of the iron-redox equilibrium in the melt. Since the endmember-partition coefficients do not vary significantly in the investigated T intervall the variation of the partition coefficient is essentially controlled by the T dependance of the redox equilibrium in the melt. The T dependance of the equilibrium constant is derived from the T variation of the Fe partitioning and from Moessbauer spectroscopic determination of ferric-ferrous ratios in quenched melts. Thus a petrogenetic grid for f(O₂) determination in water saturated systems in the T range 750°-900°C can be developed. Of course, extrapolation to systems with largely different compositions is uncertain. Therefore the applicability of the oxygen barometer was tested by measuring of the D_Fe between plagioclase and glass (inclusion & matrix) of natural volcanic samples from Mt. Pelee and Mt. Pinatubo where the preeruptive conditions are bracketed by other methods (Evans & Scaillet, 1997; Martel et al.; 1998). Comparison of the determined f(O₂)-T conditions shows perfect agreement for the preeruptive conditions of Mt. Pinatubo (Evans & Scaillet, 1997). For the Mt. Pelee a significantly higher f(O₂) (1.1 - 1.5 log units) for the T intervall given by Martel et al. (1998) is obtained. D_Fe may be influenced not only by T and f(O₂) but also by the compositions of melt and crystal and the water activity in the system. As both natural examples do not largely differ in composition, the other parameters have to account for the observed discrepancy.

Evans BW & Scaillet B, American Mineralogist, 82, 625-629, (1997).

Martel C, Pichavant M, Bourdier JL, Traineau H, Holtz F, Scaillet B, Earth Planet. Sci. Letters, 156, 89-99, (1998).

Rutherford MJ & Devine JD, Fire and Mud: eruptions and lahars of Mt. Pinatubo, PHIVOLCS and Univ. of Washington Press, Seattle, 751-766, (1996).

O08 : 4B/34 : H2

Two Types of Garnets in the Metamorphic Rocks of Sub-Polar Urals

Alexander Pystin (reg@geo.komi.ru)

Institute of Geology, 54, Pervomayskaya st., Syktyvkar, 167000, Russia

The garnetiferous metamorphic rocks of Sub-Polar Urals are found in two structural levels: Lower and Upper Pre-Cambrian. The study of the chemical composition of garnets shows that all of them refer to grossular-pirope-almandine series and are represented mainly by zonal (in chemical composition) crystals. By the type of zonation garnets are subdivided into two groups.

Garnets with progressive type of zonation where the content of magnesium increases and the content of manganese decreases from the central to peripheral parts of the grains predominate. They are known throughout the studied section.

The other less abundant group is represented by garnets with regressive type of zonation: the grains centers are enriched in magnesium and the margins - in calcium. It is known that garnets with calcium enriched outer rims are very typical of the rocks which have been multiply metamorphosed uder higher pressures and lower temperatures. Such garnets have been revealed only in the lower Early Pre-Cambrian part of the section and are absent in the overlying Riphean successions.

Zonation revealed in the garnets of the second group cannotbe connected with regressive branch of the monocycle metamorphism. If the latter is the case the grains margins have enriched manganese contents with the same amount of calcium as in crystal nuclei. The stages of metamorphism must have occurred in succession. The second stage, dealing with related to crystallization of the outer parts of grains, was reguessive in terms of temperature and progressive in pressure.

On the basis of these two types of garnets with different chemical zonation and their distribution within the section of Pre-Cambrian metamorphic rocks of Sub-Polar Urals, the following conclusions can be made:

1. Lower Pre-Cambrian rocks of the area are multiply metamorphosed: they under went not less than two stages of metamorphism in P-T stability regime of garnets of grossular-pirope-almandine composition. Upper Pre-Cambrian rocks were subjected to one stage of metamorphism in the same P-T stability regime ofgarnets.

2. The latest of these two stages of metamorphism was common for both Lower and Upper Pre-Cambrian rocks. The crystallization of garnets with progressive zonation and growth of calcium rims over the already formed garnets with the development of regressive zonation is associated with this stage.

O08 : 4B/35 : H2

Super-Silicic Garnet Microstructures from an Orogenic Garnet Peridotite; Evidence for an Ultra-Deep (>7 GPa) Origin

Herman van Roermund (hermanvr@geo.uu.nl),

Martyn Drury,

Auke Barnhoorn &

Almar de Ronde

Geogynamic Research Institut, Faculty of Earth Sciences, Utrecht University, 3508 TA, Utrecht, The Netherlands

Relicts of super-silicic garnets have been discovered for the first time within orogenic garnet peridotites, Otrøy, western gneiss region, Norway. The evidence consist of two pyroxene exsolution from garnet. Super-silicic garnets are only stable at depths greater than 150 km. Estimates of the initial composition of the super-silicic garnets imply pressures of 6-7 GPa indicating that the Otrøy garnet peridotite were derived from depths > 185 km. The Otrøy garnet peridotites consist of interlayered fertile and depleted compositions with ~ 50% lherzolite and 50% harzburgite plus dunite. Three types of garnets were recognised: 1) normal matrix type, grain-size < 4 mm, 2) isolated large single crystals (grainsize < 4 cm) and 3) isolated garnet nodules up to 10 cm in size. Garnet reveals two characteristic microstructures: i) all garnets are heavily fractured and ii) nodular garnets contain two-pyroxene exsolution microstructures. The deformation-induced garnet microstructure pre-dates the well-known two pyroxene-spinel coronas as fractures are truncated by coronas. In addition garnet plasticity is indicated by occurences of high densities of naturally decorated dislocations. Decoration is (thought to be) due to spinel precipitation. Large nodular garnets consists of mm-scale (2-8 mm) garnet grains with interstitial orthopyroxene. In the larger garnet cores two-pyroxene exsolution microstructures occur. Pyroxene exsolution is absent in an exsolution-free zone within 2 mm of garnet grain boundaries. The two-pyroxene exsolution microstructure pre-dates garnet fracturing as pyroxene needles become off-set along fractures. In addition at such microstructural sites exsolved pyroxene becomes replaced by later alteration products. Estimates of the PT conditions for mineral chemical equilibration were obtained from geo-thermometry and barometry. The mineral compositions equilibrated at mantle conditions around 805 ± 40°C and 3.2 ± 0.2 GPa. Available geochronological data suggest a multi-stage, multi-orogenic exhumation history.

O08 : 4B/36 : H2

The Alpine Metamorphic Pattern in the Danubian Window, South Carpathians (Romania)

Magda Ciulavu (ciulavu@ns.igr.ro)¹ &

Rafael Ferreiro Mählmann (ferreiro@ubaclu.unibas.ch)²

¹ Institutul Geologic al Romaniei, 1 Caransebes str., RO 78344, Bucharest 32, Romania

² Mineralogisch-Petrographisches Institut, Bernoullistrasse 30, CH-4051, Basel, Switzerland

The central part of the Southern Carpathians comprises three main nappes systems. From the lowest to the upper tectonic unit they are: 1) Lower and Upper Danubian, 2) Severin-Cosustea, and 3) Getic. The Danubian nappes are made of continental basement and Mesozoic cover, while the Severin nappe is an ophiolite-flysch unit and the Cosustea nappe is a tectonic mélange. The Danubian and Severin units have been exhumed below a low angle detachment (Getic detachment) during Eocene to Oligocene.

The samples analyzed come form the metapelitic Jurassic and marly Cretaceous rocks of the Danubian and from Cretaceous pelagic and flysch formations of the Severin and Cosustea nappes. The samples were analyzed by XRD, optical microscopy and microprobe. The XRD studies were performed in order to determine the clay mineral association and the illite crystallinity (IC). The rock maturity was determined by optical microscopy using the vitrinite reflectance method. On the basis of optical microscopy, followed by microprobe analyses, P-T conditions of the metamorphism were calculated using chlorite thermometry and phengite barometry.

In the case of the Cretaceous samples, the mineralogy of the clay fraction is made of illite + chlorite + quartz ± mica/paragonite interstratification ± hematite. The Jurassic formation shows a similar mineralogy in the south-western part of the area, while in the NE a more complex mineral association has been identified: illite + paragonite + chloritoid + pyrophyllite + quartz ± chlorite ± mica/paragonite interstratification. In this last case, because of the interference of the peaks from several minerals in the 10 Å region, the IC was determined by deconvolution or/and calculated from the width of the 2 Å peak. The illite crystallinity values show obvious trends, namely an increase of illite aggradation from SW to NE, and from S to N. In the eastern part of the area, vitrinite reflectance shows a rank of maturity of the anthracite and meta-anthracite stage; close to the Getic detachment the values increase to the semi-graphite stage. Lower maturity values are characteristic for the western part. From chemical analyses of chlorite and mica/phengite, temperatures of 340-400°C and minimum pressure of about 3 kbar have been deduced for the NE. Higher minimum pressure conditions (~5 kbar in the Danubian and 4.5 in the Severin) are inferred for an earlier (pre-detachment) mica/phengite generation.

The results indicate an increase of metamorphic grade from SW to NE and from S to N. The SW-NE trend can be explained by the exhumation of a metamorphic core complex given by the Getic detachment. The S to N gradient may be inherited from pre-detachment top to S nappe stacking.

O08 : 4B/37 : H2

Corona Formation During Prograde Metamorphism of Gabbronorites

Tatiana Larikova (tata@igem.msk.su)

IGEM, Staromonetny per, 35, Moscow, Russia

Cpx-Hbl-Grt coronas in metamorphosed labrodorite-gabbronorites of the PR massif Tolstik in N.Karelia were studied. In minimal altered rocks were found the following corona structures between primary Opx and Pl: Opx/Cpx+Hbl+Qtz/Grt+Qtz/Pl. Garnet porphiroblastes are also present in this assemblage. Two main paragenesises have been distinguished: Cpx+Grt+Qtz belongs to the corona formation, and the Hbl is secondary. The Grt porphiroblastes show particularly pronounced prograde zonation. From its core to rim, Mg/Mg+Fe significantly increases and Ca content decreases. In the central part of garnet layer there is a peak in the Ca content, and Mg/Mg+Fe increases slightly from the inner to the outer part of the corona. This similarity in the composition and the structure of the Grt grains and coronas suggests that the formation of the coronas started in the prograde stage of the metamorphic event, approximately at 670°C and 5-6 kbar. The growth of Cpx and Grt coronas in the boundary between primary Opx and Pl started according to the reaction Opx+Pl=Cpx+Grt+Qtz by mechanism of diffusion metasomatism with opposing diffusion of Ca from Pl, and Mg and Fe from Opx. The diffusion rate of Ca controls the growth and the amount of Cpx. Mg and Fe controls the Grt. Si in this process is quite inert, the diffusion rate of Al is very slow and does not cause the formation of any phase. Composition of the coronas is determined only by the chemical potentials of Fe, Ca, and Mg. The calculated Onzager coefficients from the model of steady-state diffusion are L_AlAl / L_MgMg =0.19, L_AlAl / L_FeFe =0.38, L_AlAl / L_CaCa =0.46.Hbl in the inner layer of the corona was formed during the later retrograde a mphibolization of gabbronorites according to the reactions: Opx+Cpx+Grt+Na₂O+H₂O=Hbl+Qtz; Cpx+Grt+Na₂O+H₂O=Hbl+Pl. At this stage (550°C and 4-5 kbar) corona structures were destroyed, the Grt coronas were crystallized into the separate grains with retrograde zonation; and the gabbronorites were transformed into the garnet amphibolites.

O08 : 4B/38 : H2

Scapolite Formation During Desiccation of Lower Crustal Shear Zone Rocks

Kåre Kullerud (kaarek@ibg.uit.no)

Dept. of Geology, Univ. of Tromsø, N-9037 Tromsø, Norway

Fluid-rock interaction resulted in the formation of Cl-enriched biotite (Kullerud, 1995) and Cl-enriched amphibole (Kullerud, 1996) during shear zone formation in a noritic gabbro in Lofoten, northern Norway. Metamorphic salt has also been described from the rock (Markl and Bucher, 1998, Markl et al. in press). This paper presents the occurrence of scapolite in the shear zone rock. Complex metamorphic coronas composed of amphibole, quartz, plagioclase and garnet formed along the contacts between igneous mafic minerals (CPX, OPX, BT and ILM) and igneous plagioclase. Igneous plagioclase (An_50-60) reacted to polygranular aggregates of plagioclase with a lower An-content (An_20-55). Abundant small grains (<0.1 mm in diameter) of amphibole, biotite, epidote, magnetite and kyanite formed within the aggregates of plagioclase. Corundum, quartz and Ba-rich muscovite occur as tiny intergrowths within the altered igneous plagioclase. The Cl-bearing minerals show large compositional variations. The corona forming amphibole shows Cl contents in the range 1-3 wt%, while amphibole occuring in the plagioclase matrix shows Cl contents in the range 3-5 wt%. The Cl content of scapolite ranges between 1.3 wt% and 3.8 wt% corresponding to a content between 42% and 75% of the marialite endmember.Textural evidence shows that scapolite formed during reactions between an externally derived Cl-bearing fluid and the igneous plagioclase. The sodium content of scapolite and plagioclase in contact, however, is poorly correlated. Clearly, scapolite composition was neither controlled by the composition of the igneous plagioclase that was replaced, nor by the composition of the secondary plagioclase which now occurs in contact with scapolite. It is inferred that scapolite formed in equilibrium with the fluid phase. Plagioclase, however, did generally not attain equilibrium due to the sluggishness of the cation exchange reactions between fluid and plagioclase. Al³⁺ was apparently nearly insoluble in the fluid phase, and resulted in the formation of kyanite and occasionally corundum within the plagioclase matrix during the breakdown of the An-component of the igneous plagioclase. The distribution of kyanite and corundum in the rock can be related to variations in the activity ratio a_Si4+/a_Al3+ of the local environments. The Cl-contents of amphibole and scapolite in contact are clearly correlated. This suggests that the compositions of the minerals were principally controlled by the activity of Cl of the equilibrium fluid during mineral growth. The compositional variations of scapolite and amphibole can be explained to be the result of successive desiccation of an externally derived Cl-bearing fluid phase. The first amphiboles that formed extracted preferentially H₂O from the fluid phase, leading to a successive Cl-enrichment of the fluid phase. Cl-rich amphibole and scapolite formed later in equilibrium with an evolved, Cl-enriched fluid phase.

Kullerud K, Contrib. Mineral. Petrol, 120, 42-59, (1995).

Kullerud K, Eur. J. Mineral, 8, 355-370, (1996).

Markl G, Bucher K, Nature, 391, 781-783, (1998).

Markl G, Ferry J, Bucher K, Am. J. Sci, (in press).

Session O08:4P

O08 : 4P/01 : PO

Lujavrites from the Ilimaussaq Intrusive Complex in South Greenland: Crystallization Conditions and Composition of the Late-Magmatic Fluid Phase

Holger Sommer (markl@ruf.uni-freiburg.de) &

Gregor Markl

Institut für Mineralogie, Albertstrasse 23 B, 79104-Freiburg, Germany

Geological and geochemical evidence suggests that the Ilimaussaq intrusion in the Precambrian Gardar rift province of South Greenland was emplaced as three main pulses - nepheline-augite syenite, quartz-bearing granite and agpaitic rocks. The agpaitic rocks in the Ilimaussaq Intrusion have been dated to ~1168 Ma (Rb-Sr w.r., Blaxland et al., 1976). The third, and main, pulse of peralkaline, silica undersaturated magma lead to the formation of foid-bearing syenitic rocks (sodalite foyaite, naujaite, kakortokite) and extremely fluid-enriched residual melts, the lujavrites. Field and geochemical observations indicate the establishment of two or more, largely isolated sub-magma chambers of lujavrite. Changes from green ägirin lujavrite to black arfvedsonite lujavrite record changes in fO₂, fH₂O and possibly in fHF and fHCl. Interestingly, textural relationships within various samples suggest alternation from aegirin to arvedsonite and back to aegirine stability indicating oscillating intensive parameters of fluid and/or liquid. The final stages of magmatism are marked by the emplacement of arfvedsonite lujavrite and lujavrite extremely rich in villiaumite (NaF) and naujakasite. Reaction textures at naujaite-lujavrite interfaces and the occurrence of hydrothermal veins with abundant analcime, aegirine or arfvedsonite record the expulsion of late-magmatic, relatively low-temperature fluids in the analcime stability field from the lujavrites. These textures involve formation of abundant needle-shaped crystals of arfvedsonite and/or aegirin in veins and at the interfaces growing into the naujaite. On the other hand, aegirine and arfvedsonite in the lujavrites are in many places converted into each other in areas close to naujaite xenoliths. This indicates involvement of fluids originally stored in the naujaite and it may indicate that the lujavrite intruded while the naujaite was not yet fully crystallized (despite sharp borders of naujaite xenoliths in lujavrite). In order to understand the compositional and intrinsic variability in the lujavrites and especially their fluids, new data on F-amphibole and other halogen-bearing minerals are combined with microthermometry results from fluid inclusions and geothermometry based on various amphibole-feldspar-nepheline equilibria. The halogen-rich nature of the fluids is demonstrated by the occurrence Cl-rich ussingite, villiaumite, abundant sodalite and extremely saline fluids with up to 60 wt.% NaCl equivalent.

Blaxland et al., Lithos, 9, 31-38, (1976).

O08 : 4P/02 : PO

Crystallization Conditions of the Augite Syenite, Ilimaussaq Intrusion, Greenland

Michael Marks (markl@ruf.uni-freiburg.de) &

Gregor Markl

Institut für Mineralogie, Albertstrasse 23 B, 79104 Freiburg, Germany

The Ilimaussaq intrusion is part of the late Precambrian Gardar rift province of South Greenland. The Gardar intrusive complexes fall into two categories: those involving silica-saturated rocks, such as quartz syenite and granite and those involving nepheline-bearing syenitic rocks. Only in the Ilimaussaq intrusion both rock types occur together. The main rock types of the complex are nepheline-bearing augite syenite, quartz-bearing alkali granite and nepheline-bearing syenitic rocks (agpaitic rocks named naujaite, sodalithe foyaite, lujavrite and kakortokite). Geological and geochemical evidence suggest that the intrusion was emplaced as three main pulses: 1) augite syenite(dated by Sm-Nd mineral isochrone to 1.13±0.05 Ga), 2) alkali granitic rocks, and 3) agpaitic rocks. In order to constrain the early evolution of the complex and to be able to compare (and relate?) the various rocks to each other, the augite syenite was examined in detail at various localities. The augite syenite consists of alkali feldspar, olivin, clinopyroxene, amphibole, biotite, nepheline, apatite and magnetite. It forms a shell around the southeastern, southern and western margin of the Ilimaussaq intrusion. This shell is between 10 and 500 m thick and it is commonly separated from the inner agpaitic rocks by an agpaitic border pegmatite. Xenoliths of the augite syenite are found in all other rock units of the intrusion. In addition, the augite syenite forms a horizontal sheet in the roof zone northeast of Lake Taseq. The rock is chilled against the country rocks which consist of Julianehåb granite and Gardar sandstones and volcanics. It shows increase in grain size towards its inner contacts within the complex. This textural trend is accompanied by chemical trends: From the outer to the inner part, the Fe and Mn content in olivine and clinopyroxene increases systematically from Fa₈₁ to Fa₉₄ and from Ac₄Di₅₀Hd₄₆ to Ac₄Di₄₁Hd₅₅ indicating decreasing temperatures with continuous fractional crystallization. Nepheline becomes more abundant with increasing distance from the outer contact indicating decreasing SiO₂ activity. Detailed phase equilibrium studies using the QUILF technique combined with microthermometry of fluid inclusion are used to discuss the intensive parameters (P, T, fO₂, aSiO₂) during crystallization of the augite syenite and its relation to the agpaitic rocks within the Ilimaussaq complex.

O08 : 4P/03 : PO

Geochemical Behaviour of F and Cl During Crystallization of the Nepheline-Syenitic Units of the Ilimaussaq Intrusion, Greenland

Gregor Schwinn (markl@ruf.uni-freiburg.de) &

Gregor Markl

Institut für Mineralogie, Albertstrasse 23 B, 79104-Freiburg, Germany

The Ilimaussaq Intrusion belongs to the alkaline, probably rift-related intrusions in the Gardar Province of South Greenland. The approximately 1.18 Ga old intrusion exhibits the greatest diversity of different rock types of all Gardar intrusions and consists of principally three rock suites: nepheline-bearing augite syenite, quartz-bearing alkali granite and a wide variety of foid-bearing syenitic rocks. These three main rock suites are considered to represent three pulses of different kinds of magmas. The alkaline granite is found in the highest portions of the intrusive complex. It intrudes the augite syenite which forms an early shell into which - partly separated from the interior by a border pegmatite - the later foid-syenitic to agpaitic magma intruded. This central part consists of a sequence of different agpaitic foid syenites from sodalite-foyaite at the top, via naujaite, lujavrite to kakortokite at the bottom. The genetic relationship among these various rock types is still not understood. They are considered to represent flotation cumulates (sodalite foyaites, naujaites with up to 80 vol.% sodalite), bottom cumulates (kakortokites, but unclear of which magma, maybe of the naujaitic magma?) and very fluid-rich residual, highly differentiated melt (lujavrites). Early anhydrous mineral assemblages (pyroxene, olivine, feldspars, nepheline) in the foyaite develop towards hydrous assemblages with F-bearing amphibole and with aenigmatite in foyaite, kakortokite and naujaite, and finally to the extremely volatile (F and Cl)-rich lujavrites. The early changes in melt composition are indicated by the chemical changes from low-Zr to high-Zr aegirin augite that is overgrown by F-amphibole which in turn is again overgrown by Zr-free aegirine. These various phase assemblages reflect variability in the composition of both the coexisting fluids and the coexisting melt. The early stages of the crystallization history of the central portion of the Ilimaussaq intrusive complex is deduced from phase equilibria among the early, anhydrous minerals. Fluid compositional changes in terms of fO₂, fH₂O, fHF and fHCl are critically important for the development of agpaitic rocks. Further insight into the variations in these parameters is gained by the combination of these calculations with fluid inclusion microthermometry and with calculations on halogen-bearing minerals (mainly amphiboles) involving aegirine. We present systematic data from all units of the central part of the Ilimaussaq intrusion complex.

O08 : 4P/04 : PO

The Bjerkreim-Sokndal Layered Intrusion; Norway: Study of a Sulphide-Bearing Interval at the Boundary between two Megacyclic Units

Kristine Krogh Jensen (kristine@geo.aau.dk.) &

J. Richard Wilson (jrw@geo.aau.dk)

C.F.Moellers Alle, bygn. 110, DK-8000 Aarhus C, Denmark

The Bjerkreim-Sokndal layered intrusion is situated in southwest Norway. It belongs to the Rogaland intrusive complex which was emplaced in late Proterozoic times during the Sveconorwegian orogen. The Rogaland intrusive complex consists of massif-type anorthosites, smaller charnockitic intrusions, jotunite dyke swarms and the Bjerkreim-Sokndal layered intrusion.

The intrusion consists of three lobes: the Mydland, Sokndal and Bjerkreim lobes. Only work on the Bjerkreim lobe is involved in this project. The magma chamber had an initial saucer-shape, which has been deformed into a syncline whose fold axis plunges 20-40° S. Lithologically the intrusion consists of rock-types ranging from anorthosite, leuconorite, troctolite, norite, gabbronorite, jotunite, mangerite, quartz mangerite and charnockite.

The lower part of the Bjerkreim lobe comprises a 7000 m thick Layered Series which has been divided into six mega-cyclic units (MCU 0-MCU IV). The MCU's crystallised from separate batches of magma. Within each MCU the rocks gradually become more evolved upwards, followed by a compositional reversal at the base of the overlying MCU. The Layered Series is overlain by more evolved, largely unlayered rocks (mangerites-charnokites).

This project is concerned with the boundary between MCU II and III. Massive ilmenite norites in the upper part of MCU II become modally layered upwards before sporadically developed olivine-bearing norites occur in the lower part of MCU III (zone b). The modally layered interval commonly has a pyroxenite (up to 3 m thick) developed at the base. A rust zone is usually present in the vicinity of the pyroxenite. The rust zone is a result of the weathering of an interval with a high concentration of disseminated sulphides. The pyroxenite occurs in discontinuous bodies and where the pyroxenite is lacking the sulphides are situated in the norites.

Fieldwork was carried out in July '98, and a series of sample profiles have been collected across the MCU II-III boundary, with emphasis on the sulphide-bearing rocks and the pyroxenite. The aim of the project is to identify the sulphide phases, to explain how the increased concentrations of the sulphides were formed, and to obtain an understanding of the magma chamber processes responsible for crystallisation of the pyroxenite and sulphides near the boundary between two MCU's.

O08 : 4P/05 : PO

Extreme Assimilation in a Crustal Magma Chamber: Sr and Nd Isotopic Constraints from the Hasvik Layered Intrusion, Norway

Christian Tegner (christian.tegner@geo.aau.dk)¹ &

Brian Robins (brian.robins@geol.uib.no)²

¹ Dep. Earth Sciences, University of Aarhus, DK-8000 Aarhus C, Denmark

² Dep. of Geology, University of Bergen, N-5007 Bergen, Norway

Petrographic, mineralogical, whole-rock major- and trace-element compositions, and Sr-Nd isotopic data for mafic cumulates, chilled margins, and country rocks of the ~700 Ma Hasvik Layered Intrusion, North Norwegian Caledonides, constrain coupled crustal assimilation and fractional crystallisation (AFC) of mantle-derived tholeiitic basalt emplaced into metasediments at a deep crustal level (6-8 kbar, 4-600^oC). Correlated variations of whole-rock ⁸⁷Sr/⁸⁶Sr (0.7038-0.7089), <epsilon>Nd (+4.8 to -3.3) and mineral compositions (e.g. X_An = 0.72-0.52, and mg#_opx = 0.77-0.39) suggest steady-state AFC with a constant ratio, r, for the rate of assimilation to the rate of crystallisation of ~0.32 for ~1100 m of cumulates demonstrating a continuous tholeiitic fractionation trend (olivine-melt reaction, Fe-Ti oxides in, apatite and pigeonite in), following a ~335 m thick cumulate sequence with a reverse cryptic variation recording continuous magma emplacement at the chamber floor. This value of r is close to the upper limit permitted by the heat budget, placing the Hasvik Layered Intrusion among the most contaminated layered intrusions presently known. Slabs of highly-modified metasediment xenoliths are common in the cumulates, particularly in the upper part of the Layered Series, and are the remnants of the material assimilated.

O08 : 4P/06 : PO

Petrological, Mineralogical and Geochemical Outlines of the Late Permian to Triassic Karavanke Granitic Belt (Slovenia)

Meta Bole (meta.bole@uni-lj.si)¹,

Tadej Dolenec (tadej.dolenec@ntfgeo.uni-lj.si)¹,

Nina Zupancic (nina.zupancic@uni-lj.si)¹ &

Breda Cinc-Juhant (breda.cinc-juhant@uni-lj.si)²

¹ University of Ljubljana, Department of Geology, Askerceva 12, 1000 Ljubljana, Slovenia

² Slovene Museum of Natural History, Precernova 20, 1001 Ljubljana, Slovenia

The rocks of the Karavanke Granitic Belt belong to a bimodal magmatic association consisting of predominant syenogranite and syenite and contemporaneous mafic and intermediate rocks of alkaline character. Mafic rocks consist of olivine-bearing gabbro and monzogabbro, wich represent about 20% of the whole massif, and range in size from decimetric microgranular enclaves to large decametric bodies. The intermediate rock types are monzodiorite and monzonite wich show field, textural and chemical features suggest that they formed as a result of the interaction between felsic and mafic magmas. The porphiry syenite with rapakivi texture, which occurs in close spatial association with mafic enclaves, is interpreted as a piece of evidence for such interaction.

Field, petrographic and geochemical evidence indicate that the rocks of Karavanke Granitic Belt do not represent the result of a crystal fractionation process, nor of a simple bulk mixing process. Only some of the more mafic and, at the other side some of the more felsic rocks could be consistently linked by crystal fractionation. Rocks of intermediate composition are the results of a different degree of interaction between the two end-member magmas.

Chemical and mineralogical compositions indicate that the basic rocks represent the mantle derived magma. Rising and fractional crystallization of basic magma probably produced the heat necessary to trigger the melting of crustal material, which caused the formation of the felsic magma, as suggested by initial Sr isotopic data (Dolenec, 1994).

The highly contrasted acid-basic relationships and the alcaline character of rocks fit an anorogenic to post orogenic tectonic setting (Pitcher, 1993) and is consistent with the regime of incipient rifting suggested by Bonin et al. (1987) for the Western Mediterranean Magmatic Province, thus indicating that the late Permian to Triassic Karavanke Granite Belt is also part of the same magmatic province.

Bonin B, The enclaves of alkaline anorogenic granites: an overview. in Enclaves and Granite Petrology, J. Didier & B. Barbarin, Elsevier, Amsterdam, 179-188, (1991).

Dolenec T, Novi izotopski in radiometricni podatki o pohorskih magmatskih kamninah, Rudarsko-metalurgki zbornik, 41, 147-152, (1994).

Pitcher WS, The Nature and Origin of Granite. Blackie A & P, London, 321, (1993).

O08 : 4P/07 : PO

Geochemistry of Polycyclical Metamorphic Rocks of the Bragança Massif (Variscan Chain, NE Portugal)

José F. Santos (jfsantos@geo.ua.pt)¹ &

José M. Munhá (jmunha@fc.ul.pt)²

¹ Departamento de Geociências, Universidade de Aveiro, 3810 Aveiro, Portugal

² Departamento de Geologia, Faculdade de Ciências, Universidade de Lisboa, 1700 Lisboa, Portugal

The most striking geological feature of NE Portugal is the presence of several exotic tectonic units emplaced upon the Central-Iberian autochthon. The upper allochthonous thrust complex crops out in the Bragança Massif, where four main rock types can be recognized: (1) high-pressure granulites; (2) eclogites (enclosed by paragneisses); (3) plutonic gabbroic bodies; (4) ultramafic rocks. Previous studies have shown that the first two lithotypes underwent a complex polycyclical tectonometamorphic evolution. Recently, in a type (1) rock, the HP/HT metamorphic event was dated as Grenvillian (1.0 to 1.1 Ga), but, most commonly, HP granulites are strongly affected by Hercynian metamorphism, giving rise to the Palaeozoic isotopic ages usually obtained, even for the polycyclical lithoptypes (granulites and eclogites).

The compositions of the Grenvillian high-pressure granulites seem to reflect fractionation processes under high-grade metamorphic conditions; redistribution of highly incompatible trace elements would have occurred, at least locally. The trace element contents of different types of granulites (from mafic to felsic), when compared to the compositions usually assumed to represent the lower crust, display LREE-depleted patterns, suggesting that, besides being remobilized, the most incompatible elements have suffered removal, probably via a silicate liquid phase.

Three main compositional groups of eclogites were defined, using immobile hygromagmatophile trace elements: A) rocks with NMORB-like compositions; B) subalkaline (tholeiitic to calc-alkaline) rocks displaying geochemical signatures typical of supra-subduction settings; C) rocks related to alkaline basalt protoliths. The diversity revealed by the eclogites together with the geochemical features of the enclosing paragneisses (characteristic of sediments from basins related to volcanic arcs) suggest that the protoliths of the metaigneous and metasedimentary rocks were formed in a back-arc marginal basin, operating during Precambrian times, most probably in the Cadomian tectonic cycle.

O08 : 4P/08 : PO

Gran Canaria: The Subvolcanics of the Roque Nublo-Cycle; Geochemical and Petrological Relations to the Extrusives

Eric B. Barnert

(barnert@hrz1.hrz.tu-darmstadt.de)¹,

Gerhard P. Brey (brey@em.uni-frankfurt.de)² &

Wolfgang F. Müller

(wmueller@hrz1.hrz.tu-darmstadt.de)¹

¹ Technische Universität, Institut für Mineralogie, Schnittspahnstr. 9, 64287 Darmstadt, Germany

² J. W. Goethe-Universität, Institut für Mineralogie, Senckenberganlage 28, 60054 Frankfurt, Germany

Alcaline subvolcanic intrusives and hauyne bearing extrusives in the centre of Gran Canaria are the main components of mono- and polymict breccia-sheets. The max. 20 m thick sheets are dominated by the hauyne-bearing extrusives with a tephri-phonolitic to phonolitic bulk composition. These rocks are higher differentiated than the subvolcanics with a tephritic to phono-tephritic character. The breccias represent a separate stratigraphic unit between the underlying Los Listos Formation and the overlying large debris avalanche deposits of the Ayacata Formation. They can be used as marker for the reconstruction of the pre-Ayacata Formation paleomorphology. A new mapping in the Roque Nublo area revealed more subvolcanics as described before. They occur as fragments in the debris avalanche deposits and therefore are no longer interpreted as subvolcanics in a primary intrusive position (Brey und Schmincke, 1980). In the monomict sheets the matrix and the components have the same phase composition as shown by powder diffraction. The secondary transport from a higher position is an argument for the theory that Gran Canaria had a higher altitude during the pliocene volcanic cycle. The recent subvolcanic deposits are between 1230 and 1720 m high - the highest point is 1949 m above sea-level. Another argument is the plutonic grain size of the subvolcanics (essexites, theralithes). This indicates that the subvolcanics were overlain by several 100 m of other volcanic material. RFA-data of 8 subvolcanic rocks and 162 extrusive rocks out of the Roque Nublo-Cycle were compared using spider diagrams following the plotting order for oceanic crust suggested by Hoffmann (1988). The subvolcanics have nearly the same composition as lavas of the upper Mesa de Junquillo- to the Los Listos Formation. The only exception is the subvolcanic rock from Las Lagunetas with a basaltic composition. It is equivalent to lavas of the lower Mesa de Junquillo Formation (lower part of the pliocene Roque Nublo cycle) and does not belong to the miocene cycle. Volatile transport and alteration causes differences in the concentration of mobile elements between the subvolcanics and their corresponding lavas. Our studies have shown that the large debris avalanche deposits of the Ayacata Formation were not formed by a single event. Many striations in a small area with different directions and between the generations of debris avalanches are an argument for a longer interval of inactivity and erosion between the events.

Brey G & Schmincke HU, Bull. Volcanol., 43-1, 15-33, (1980).

Hoffmann AW, Earth Planet Sci. Lett, 90, 297-314, (1988).

O08 : 4P/09 : PO

Geochemistry and Liquid Line Descent of the Tarrafal Nephelinitic and Basanitic Lavas, Santo Antão, Cape Verde Islands

Birgitte Printz Christensen (cb260874@geo.geol.ku.dk) &

Paul Martin Holm (paulmh@geo.geol.ku.dk)

Department of Petrology, Geological Institute, University of Copenhagen, Øster Voldgade 10, DK-1350 København, Danmark

Cape Verde Islands are situated in the Atlantic Ocean about 500 km west of Africa between 15° and 17°N and 22° and 26°W. The archipelago consists of ten major islands and several smaller islets and can be grouped into a northern and a southern array, where Santo Antão is the westernmost located island in the northern array. The islands are related to an active hotspot below the slowly moving African plate. The island of Santo Antão offers very good opportunity to study the evolution of nephelinitic and basanitic melts, as the entire 779 km² island is made up from suites of rocks derived from such magmas. From the village of Tarrafal in the western part of Santo Antão a deeply eroded valley trends SW-NE. In this valley a stratigraphic section of 500 m of accessible volcanics are exposed. The volcanics consist mainly of pahoehoe type lavas. Very few pyroclastic deposits of lapilli and bombs and a sequence of two ignimbrites are also observed. The rocks are all silica undersaturated and can be grouped into picrobasalts, basanites, tephrites, melanephelinites and nephelinites. The major phenocryst phases in the lavas are olivine, clinopyroxene and Fe-Ti oxides, whereas plagioclase and haüyne occur only in minor amounts. Haüyne occurs only in the most evolved tephrites with a MgO content of around 5 wt%. Some of the lavas are highly porphyritic with up to 40 vol% phenocrysts. With decreasing MgO content from 18.6 to 4.7 wt%; SiO₂, TiO₂, Al₂O₃, FeO_total, K₂O, P₂O₅ and MnO increase, while CaO decrease slightly. The magmatic evolution is unrelated to stratigraphic position. A simple general differentiation is indicated involving early fractionation of olivine ±chromite, which is joined by clinopyroxene at 10 wt% MgO. The olivine fractionation is evident from the negative correlation of MgO vs. Ni, over the whole range of MgO contents. Moreover Cr and Co contens decrease with decreasing Ni. Sc and CaO contents decrease for a Ni content of less than 200 ppm, which indicate the onset of clinopyroxene fractionation. Fractionation to more evolved phonolitic compositions as observed in other areas of Santo Antão is not evident from the volcanics of this profile. The variations of MgO, Ni and Cr vs. stratigraphic height are similar and several samples with high MgO, Ni and Cr content occur up through the profile. Variation diagrams, especially the ones concerning incompatible elements show more than one fractionation trend with MgO, which indicate that not all lavas are comagmatic. The primitive character of the volcanics, the variation diagrams and the stratigraphic variation of MgO, Ni and Cr could suggest an evolution controlled by repeated injections of primitve magma into a fractionating magma chamber.

O08 : 4P/10 : PO

Geochemical Data from Two Lava Sequences, Santo Antão, the Cape Verde Islands

K. Malene Hein (hm021271@geo.geol.ku.dk) &

Paul Martin Holm (paulmh@geo.geol.ku.dk)

Department of Petrology, Geological Institute, University of Copenhagen, Øster Voldgade 10, DK-1350 Københanv, Danmark

The Cape Verde Islands are situated in the E. Atlantic Ocean between 15°-17°N and 22°-25°W. There are nine major islands arranged in a northern and southern archipelago. They overlie the uplifted ocean floor of the Cape Verde Rise. Active volcanoes are found on the two westernmost islands in the southern archipelago. The formation of the Cape Verde Islands is believed to be related to an underlying mantle plume. Santo Antno is the western most island in the northern archipelago. It has an area of approxi-mately 780 km² and a rugged topography reaching about 1930 m asl. The volcanics on the island are silica-undersaturated which is a characteristic feature of all the Cape Verde Islands.

We present data from two profiles on Santo Antno; the Agua Nova profile in the northwestern part and the Escabecada profile in the southwest. The profiles consist chiefly of lava flows with few dykes. The rock types range from alkaline basalts (ankaramites) through tephrite to haüyne or nepheline porphyritic phonolitic tephrite and tephriphonolite.

Magmatic differentiation due to crystal fractionation is reflected in the geochemistry of the lavas. Ni decreases as MgO decreases from 18 to 2 wt%. This is evidence of fractional crystallisation of olivine through the entire sequence. Sc and CaO decrease markedly from around 9 wt% MgO reflecting fractionation of clinopyroxene. From 6 to 2 wt% MgO, the SiO₂ content increases significantly while TiO₂ and to a lesser degree P₂O₅ decrease demonstrating the removal of Fe-Ti oxide and apatite. The content of Al₂O₃ increases with decreasing MgO indicating that Al-bearing phases like plagioclase, amphibole and feldspatoids have no significance as fractionating phases.

Generally, the Agua Nova samples have lower SiO₂, Al₂O₃ and Ni content and higher content of TiO₂, P₂O₅ and incompatible elements compared to the Escabecada samples. This is evidence that different magmas were erupted. The higher content of incompatible elements point to a smaller degree of melting of the Agua Nova source or a source that is more enriched. Additionally, the range in composition for a given MgO content in either of the profiles excludes the possibility of the volcanic products within a single profile being comagmatic.

In the Escabecada profile the stratigraphic variation of MgO seems to reveal two pulses of primitive magma with differing contents of SiO₂. With these two primitive magmas the diversity in SiO₂ of the more primitive lavas may be explained. A development towards more evolved compositions is seen towards the top. In the Agua Nova profile no clear development is seen with stratigraphic height.

O08 : 4P/11 : PO

Petrology and Geochronology of Santo Antão, Cape Verde Islands ­ A Preliminary Report

S. Plesner (SUSANNE@geo1.aau.dk) &

J. R. Wilson (JRW@geo1.aau.dk)

Department of Earth Sciences, Geological Institute, University of Aarhus, 8000 Århus C, Denmark

The Cape Verde Islands are located in the Atlantic, 15-17°N and 500 km west of Senegal. Santo Antão (780 km²), the westernmost island of the northern chain, has not been volcanically active in historic times. The ages below were obtained in 1998 from ⁴⁰Ar/³⁹Ar incremental heating experiments at Oregon State University.

A ca. 800 m-thick sequence (Chã de Morte Series) has been sampled in the central part of the island (Fig.1). The lowest exposed lavas are basalts/picrobasalts with an age of 5.61±0.13 Ma. These are overlain by ca. 220 m of more evolved lavas. These varied lavas are intercalated with (picro)basalts through the next ca.320 m. Towards the top of the Series the lavas become more evolved, reaching phonolites at 1400 m. Tephrites at the very top of the profile have an age of 1.20±0.01 Ma. The Chã de Morte Series is cut by numerous dykes. The variations in lava-type and MgO-contents indicate repeated events of magma chamber development.

Lithology, whole rock MgO% and ⁴⁰Ar-³⁹Ar ages of the Chã de Morte Series in the central part of Santo Antao. Rock names are based on a TAS-diagram. Heights are in meters above sea level.

Much of the island is covered by lapilli deposits and minor lava flows erupted from locally well-preserved volcanic cones. These deposits are post-dated by a phonolitic pumice (the Cão Grande pumice) that drapes over much of the present topography. The vent from which it was erupted has been buried by lapilli deposits and minor lava flows in the Tope de Coroa area (0.20±0.01 Ma). The Cão Grande pumice has an age between 0.43 and 0.20 Ma. The youngest volcanic activity is represented by nephelinitic lavas near Porto Novo in the south of the island (0.08±0.01 Ma).

The youngest volcanic activity on Maio, the oldest island in the southern chain, took place 7 Ma years ago. If the islands become progressively younger towards the west, the presence of >5.61±0.13 Ma old lavas on Santo Antão does not leave much time for development of the intermediate islands unless some of them were formed simultaneously.

O08 : 4P/12 : PO

Dykes and Sills of Santo Antão, Cape Verde Islands

Lisbeth Hansen (lisbeth@geo.aau.dk) &

J. Richard Wilson (jrw@geo.aau.dk)

Department of Earth Sciences, University of Aarhus, 8000 Aarhus C, Denmark

Santo Antão belongs to the Cape Verde archipelago that comprises a group of volcanic islands situated in the Atlantic Ocean 15-17° north of the equator and 300-600 kilometres west of Senegal.

Dykes and sills are abundant on Santo Antão; the purpose of the present project is to study their orientations, relative ages and distribution in different parts of the island and establish their petrographic and geochemical variation. The proportion of dykes compared to sills is large and in some areas sills are rare. Large parts of the island are covered by recent pyroclastic deposits, which hide the older lava sequences and dykes/sills. Most dykes are 40-200 cm wide and they commonly have chilled margins against the host lavas. Since they usually are more resistant to weathering than the lavas, many of the dykes stand out like isolated walls in the landscape. Some dykes have been observed to have acted as feeders to local volcanoes preserved on the tops of recently eroded steep mountainsides.

In some areas the abundance of dykes is much greater than elsewhere, indicating that dyke activity was most intensive there. The abundance of dykes in these areas implies considerable dilation; a rough estimate based on the abundance and width of the dykes is 10-20%. On the basis of orientations, crosscutting relationships and macroscopic compositions at least five generations of dykes have been recognised in the field.

There are large, systematic variations in the dominant strike of the dykes between the regions, but chemical analyses do not reveal significant geochemical differences. The majority of the dykes have 40-45% SiO₂ (volatile-free basis) and 3-7% Na₂O + K₂O, plotting as nephelinites and basanites/tephrites in a TAS diagram. The samples that fall in these fields are classified according to the "norm ne versus norm ab classification" suggested by Le Bas (1989). This results in more than half of the dykes that are basanites/tephrites according to the TAS-diagram being classified as either nephelinites or melanephelinites. Some dykes have more evolved compositions, plotting as phonotephrites and tephriphonolites, whereas a few are less alkaline and evolve towards trachytes. The most primitive dykes (7-16% MgO) plot on an olivine control line, whereas the more evolved ones (1-7% MgO) have decreasing CaO and TiO₂, consistent with the fractionation of clinopyroxene and Fe-Ti oxide. Other minor phases may also be involved. These compositions are consistent with those of the lava sequences and younger volcanics, in keeping with the dykes having acted as feeders for these extrusives.

Le Bas MJ, J. Petrology, 30(5), 1299-1312, (1989).

O08 : 4P/13 : PO

Recent Volcanism on Santo Antão, Cape Verde Islands

Rikke Pedersen (rikkep@geo.aau.dk),

Anette K. Mortensen &

J. Richard Wilson

Hasle Centervej 239,3.tv, 8210 Århus V, Denmark

The Cape Verde Islands are situated 500 km W of Senegal between 15-17°N of the equator. The Cape Verde Islands consist of two chains, where Santo Antão (SA) is the youngest and most westerly island of the northern chain. This abstract presents a short description of two recent volcanic events on Santo Antão, the Cão Grande phonolitic pumice deposit and the Porto Novo nephelinitic lava flows.

The pumice occurs as a white, layer, which is patchilly preserved over the entire island, which has an area of 780 km². The deposit is thickest (up to 8 m) and most widespread in the southwestern and central parts of SA. Over much of SA the Cão Grande pumice represents the youngest volcanic deposit and elsewhere it provides a valuable marker horizon. In the central and western parts of SA it is locally covered by lapilli deposits and minor lava flows. In the SE part of the island it is overlain by the Porto Novo nephelinitic lava flows which represent the youngest volcanic activity on the island.

These two volcanic events represent end-members of SiO₂-undersaturated alkaline magmatism. The phonolitic pumice represents an evolved magma resulting from extensive fractionation in a magma chamber, whereas the primitive, nephelinitic lava, which contains numerous mantle nodules, has come more or less directly from a mantle source.

The vent from which the Cão Grande pumice was erupted has not been found; it appears to have been buried by younger lapilli deposits. The approximate location has, however, been estimated by mapping the variations in thickness of the deposit and the size distribution of lithic- and pumice fragments. Isopleth and isopach maps indicate that the volcanic vent must have been in the western part of SA.

The Porto Novo lava was extruded from several vents in the central part of SA and flowed to the SE, reaching the coast; the flows cover an area of ca. 40 km². The sequence consists of at least 7 flows separated by either thin scoria deposits or lapilli layers. The absence of laterite horizons indicates that the flows were extruded within a short time span.

Based on ⁴⁰Ar-³⁹Ar dating carried out by Susanne Plesner, it is now possible to determine that the eruption of the Cão Grande pumice must have taken place at some time between 430 and 200 Ky ago and the Porto Novo lavas were erupted about 80 Ky ago.

O08 : 4P/14 : PO

Eruptive History of Mt. Erciyes Stratovolcano (Central Anatolia, Turkey)

Erdal Sen (sen@jeo.hun.edu.tr),

Biltan Kurkcuoglu (biltan@jeo.hun.edu.tr) &

Erkan Aydar (erkan@jeo.hun.edu.tr)

Hacettepe University, Dept. of Geological Engineering, Beytepe- Ankara, Turkey

Mount Erciyes is the most voluminous volcano (3917 m) in the Central Anatolia. The evolutionary stage of Mt. Erciyes is divided into two stages: Koç Dag stage; The volcano was mainly built up by the lava pile. The known first lavas correspond to the alkali basaltic flows. The dominant nature of lavas is characterised by andesitic lava pile. The monogenetic vents as cinder cones are basaltic andesitic in composition that represent the weak explosive activities. The lava generation was followed by pyroclastic eruptions in two sequences which led to the caldera collapse (18 x 14 km) 2.8 Ma ago. The first sequence witnessed a large episode of plinian phases with four eruption units and pumice flows which were over 30 km from the caldera. The emplacement of Valibaba Tepe ignimbrite occurences was defined as secondary. It mostly represents well-welded rheology with 3700 km² of areal distribution and is the LAR type ignimbrite. This stage was terminated by andesitic and dasitic dome extrusions which used the caldera boundary.

Recent Erciyes stage; Two phases were distinguished in its evolutionary history. First phase is characterised by effusive and weak extrusives activities which are andesitic lava flows, dasitic dome flows, basaltic andesitic lava flows, dasitic domes and andesitic cones. Second phase witnessed the dasitic explosive activities, which were well concentrated on the summit area, producing the block and ash flows, and the rhyodacitic activities that were related to dome emplacements of Dikkartin and Perikartin. Prior to these dome extrusions the plinian air fall products, surges and pumice flows were produced during this period. The last volcanic activity is represented by the debris avalanches deposits related to the summit of the volcano forming an avalanche caldera.

O08 : 4P/15 : PO

Volcanological Evolution of Kula Volcanics, Western Anatolia, Turkey

Erdal Sen (sen@jeo.hun.edu.tr)¹,

Erkan Aydar (erkan@jeo.hun.edu.tr)¹,

Hasan Bayhan (bayhan@jeo.hun.edu.tr)² &

Alain Gourgaud

(gourgaud@opgc.univ-bpclermont.fr)²

¹ Hacettepe University, Dept. of Geological Engineering, Beytepe- Ankara, Turkey

² Universite Blaise Pascal, Laboratoire de Geologie, 5, rue Kessler, 63038 Clermont-Ferrand Cedex, France

Detailed geologic mapping of Kula alkali basalts show that at least 70 seperate eruptions have occured from 1.1 Ma to historical times covering about 1000 km² in a tectonic depression. Areal coverage, physical properties and age of the products led to us divide the basaltic volcanism into three generations: First, second, and the third (FG, SG, and TG). Their eruptions were accomponied by formation of NE-SW and NW-SE oriented grabens under extentional stress in Western Anatolia. Some eruptive vents were offset from the graben axis.

FG basalts (1.1 Ma) probably erupted from an elliptical fracture zone. They crop out to eastern side of study area and form lava plateaus with a very well devoloped columnlar joint structures. In contrast to FG; SG (0.3 Ma) and TG (10000 years) alkali basalts resulted in primarily monogenetic features: cinder cones, spatter cones, and maars with lava flows and base surge deposits. Explosive activitiys of maars were replaced by dry eruptions changing vent position to cinder cones on the rim of the crater. The products of individual eruptive centers vary in distribution from 1 km² to 17 km². While FG and SG lava flows are massive, TG basalts have a porous and black appearence having the flow features of aa lavas.

Although petrological data indicate similarities between different vents, their phenocrysts assemblage can be different such as for FG: Leucite (Lc)-olivine (Ol)-clinopyroxene (Cpx) and for SG, TG: Lc-Ol-Cpx-hornblende (Hb); Ol-Cpx-Hb; and Cpx-Hb. SG and TG basalts are rich in enclaves, which are harsburgite, pyroxene-hornblendite, hornblend-pyroxenite, gabbro, and metamorphic rocks. Hornblendites and pyroxenites (5-30 cm in size) have vesicules and glass including microlites. Olivine bearing xenolithes are only present in the maar deposits. Textural and mineralogical evidences show that enclaves were related to a zoned magma chamber and were erupted in a short time.

O08 : 4P/16 : PO

Characterization of The Miocene Acidic Sequence of The Pire Mahuida Volcanic Field, Extraandean Patagonia, Argentina

Flavia Salani (fms@gl.fcen.uba.ar)

CONICET-Dto.Cs. Geologicas. Universidad de Buenos Aires, Pab. II. Ciudad Universitaria, Buenos Aires, Argentina

The effusion of a Miocene acidic volcanism is exposed at the Pire Mahuida Sierra (68° 30' W Long.; 42° S Lat.) located in the southern part of Argentina, in the Extraandean Patagonia region. It belongs to the Pire Mahuida Volcanic Field, a complex of 800 km² and a volume of 100 km³, constituted by rhyolitic and dacitic lavas, and pyroclastics of rhyolitic composition, related to lesser amount of basaltic flows. The suite represents a back-arck assemblage developed in an extensional regime. The objetive of this work is focused on the characterization of the acidic rocks through stratigraphic, morphological and petrogenetic analysis. The rhyolitic lavas occur as eight aligned domes distributed with a subcircular pattern, defining the western boundary of the volcanic field, and represent the first eruptive event. They show flow banding, vesiculation textures, litophysaes, and often, large gas cavities. In the central area of the sierra the rhyolites crop out as less viscous flows, and they correspond to a second eruptive stage. Most lavas are aphyric or poorly porphyritic, with sanidine, and quartz as phenocryst, and as intergrow in the groundmass. The dome lavas contain biotite, while the flows of central region show little amount of arfvedsonite. Trydimite is the common mineral of the vapor phase crystallization. The dacites occur in the southwestern part of the field as porphyric hornblenditic lava domes. A pyroclastic sequence dominates the northern part of the region. It corresponds to the main explosive event represented by ignimbrite sheets, surges and pyroclastic flows associated to rhyolitic domes. They show high proportion of vitroclasts, and subordinated crystals of sanidine, quartz, and biotite, and volcanic lithoclasts of jurassic andesites. Small volume of younger pyroclastic rocks, are related to lava-pyroclastic cones, and crops out in the central area. They contain vitroclasts, sanidine crystals, alkaline pyroxenes, and in some cases olivine and scarce lithoclasts. The petrographic and chemical features indicate two evolutive patterns: 1) a calcakaline high silica rhyolite-dacite sequence of lava domes, and pyroclastics of northern area; 2) a rhyolitic suite (lava and pyroclastic flows) with alkaline affinities belonging to the central region. The proposed model for the generation of the acidic magmatism is partial crust melting, where the heat for the fusion is supplied by the intrusion of basaltic magmas in an extensional regime. In the further evolution of the liquids, participate different processes such as fractional crystallization and volatile accomplishment.

O08 : 4P/17 : PO

Upper Mantle and Lower Crustal Phenomena During Intraplate Magmatism at the Onset of the Variscan Cycle: Geochemical Evidence from the Bragança Massif (Galicia-Trás-Os-Montes Zone, NE Portugal)

José F. Santos (jfsantos@geo.ua.pt)¹,

José M. Munhá (jmunha@fc.ul.pt)²,

Fernando O. Marques (gfmogm@fc.ul.pt)² &

Graeme Rogers (grogers@geology.gla.ac.uk)³

¹ Departamento de Geociências, Universidade de Aveiro, 3810 Aveiro, Portugal

² Departamento de Geologia, Faculdade de Ciências, Universidade de Lisboa, 1700 Lisboa, Portugal

³ Isotope