Origin of oceanic phonolites by crystal fractionation and the problem of the Daly gap: an example from Rarotonga

Felsic alkalic rocks are a minor component of many ocean island volcanic suites, and include trachyte and phonolite as well as various types of alkaline and peralkaline rhyolite. However, there is considerable debate on the nature of their formation; for example, are they formed by partial melting of anomalous mantle or the final products of fractional crystallization of mafic magmas. The phonolites and foidal phonolites on Rarotonga were formed by low pressure crystal fractionation of two chemically distinct parental magmas. Low silica and high silica mafic magmas produced a basanite-foidal phonolite series and an alkali basalt-phonolite series, respectively. The foidal phonolite composition evolved from the low silica mafic magmas by approximately 60% fractionation of titanaugite + leucite + nepheline + magnetite + apatite. Fractionation continued with the crystallization of aegirine-augite + nepheline + kaersutite + magnetite + apatite. The phonolites formed from the alkali basalts by approximately 40% fractionation of kaersutite + titanaugite + Fe-Ti oxide + plagioclase + apatite and continued to evolve further by fractionation of anorthoclase + nepheline + aegerine-augite + Fe-Ti oxides. As the magmas fractionated in both suites, their overall viscosities (solid + liquid) increased until a point was reached whereby viscosity inhibited the eruption of magmas with compositions intermediate between the mafic rocks and the felsic rocks. However, the magmas continued to fractionate under static conditions with the residual fluid becoming foidal phonolitic in the low silica suite or phonolitic in the high silica suite. These phonolitic liquids, as a result of an increase in volatiles and enrichment of alkalis over aluminum, would actually have a lower viscosity than the intermediate liquids. This decrease in viscosity and the switch from a magma chamber being predominantly a liquid with suspended solids to a solid crystalline network with an interstitial liquid enabled phonolitic liquids to migrate, pool, and eventually erupt on the surface.     

Basic lavas of the Archean La Grande Greenstone belt: Products of polybaric fractionation and crustal contamination

Despite the fact that some greenstone belts preserve the record of contemporaneous komatiitic and tholeiitic volcanism, a genetic link between the two is not widely accepted. The significance of a compositional gap seperating these magma types and differences in their respective degree of light rare earth element (LREE) enrichment, cited as evidence against a derivative relationship, are complicated by the possibility of crustal assimilation by magmas of komatiitic affinity. In the Archean La Grande Greenstone belt of northern Quebec a succession of metamorphosed tholeiitic basalts and younger, high-Mg, LREE-enriched andesites are preserved. The tholeiites are differentiated basaltic rocks whose chemical compositions appear to have been controlled by low pressure, gabbroic fractional crystallization and are similar to Type 1 MORB. Parental magmas were probably high-Mg liquids of compositions similar to komatiitic basalts which also occur in the greenstone belt. These high-Mg liquids are believed to be themselves the product of high pressure, OLIV+OPX fractional crystallization of more magnesian primary liquids of komatiitic composition. The higher La/Sm ratios of komatiitic basalts and tholeiites relative to komatiites in this belt, can be explained by small degrees of crustal assimilation. In the central part of the belt, late-stage, mafic igneous rocks have chemical compositions similar to Archean examples of contaminated volcanic rocks (e.g., Kambalda, Australia). These late-stage lavas consist of basalts and andesites with high-Mg, Ni and Cr abundances, LREE-enriched profiles and low Ti abundances. They are believed to be the products of crustal assimilation and crystallization of OPX-PLAG-CPX from high-Mg liquids of komatiitic affinity. The volcanic stratigraphy records the progressive effects of crustal contamination through time. A light sialic crust may have initially acted as a density barrier, preventing the eruption of primary high-Mg liquids and forcing fractionation at depth which produced more buoyant compositions. With subsequent thinning of the crust, the density barrier presumably failed, and primary liquids migrated directly toward the surface. Reaction of these liquids with tonalitic crust produced contaminated differentiates.     

Mass change calculations in altered rock series

A technique is presented whereby precursors of altered rocks in a fractionated rock series can be identified, and changes in chemical components, mass and volume accurately calculated. The procedure utilizes elements with high degrees of immobility in hydrothermal and associated interactive water-rock systems. Igneous rock series, and sedimentary and metamorphic rocks with primary continuity of chemical composition, are amenable to this treatment. In the procedure for igneous rocks, best-fit fractionation lines are established for an immobile igneous compatible-incompatible element pair, the latter element functioning as the fractionation monitor. Alteration paths are linear and pass through the origin. Intersections of the fractionation and alteration lines yield the immobile element concentrations in the precursors to the altered rocks; mobile components are determined from fractionation lines and monitor values. Total mass changes are proportional to the displacement of altered samples from the fractionation line. Masses of other elements in an altered sample (reconstructed values) are recalculated to the concentration of the monitor element in the precursor. Chemical changes are the differences between the calculated precursor compositions and the reconstructed altered rock compositions. Errors are mainly a function of the positioning of the fractionation lines.     

A potassium-rich Alkalic Suite from the Deccan Traps,Rajpipla, India

The Rajpipla Alkalic Suite is the most potassium-enriched group of basaltic rocks so far described from the Deccan Traps. In the same area however early tholeiitic flows and late tholeiitic dykes show the potassium-poor nature characteristic of most Deccan Trap magmas. The rocks of the alkalic suite are highly porphyritic and their major element variation can be interpreted in terms of crystal fractionation dominated by clinopyroxene. Plagioclase, which is an important phenocryst phase, has fractionated only in relatively small amounts as a result of a lack of density contrast between it and the liquids. A dyke-like form for the magma chambers in which fractionation has taken place is postulated to account for the abundance of highly porphyritic types. The Rajpipla area is also notable as being one of the few Deccan localities where rhyolites are found.Abbreviations AB ankaramitic basalt - PB porphyritic basalt - PTB porphyritic trachybasalt - FPM feldsparphyric mugearite - M mugearite - TR trachyte - P. RHY potassic rhyolite - Th. B. tholeiitic basalt - Th. D. tholeiitic dolerite - Af alkali feldspar     

Picritic Melts in Kilauea--Evidence from the 1967-1968 Halemaumau and Hiiaka Eruptions

Lavas from the 1967 Halemaumau eruption and picrites eruptedin August 1968 at Hiiaka Crater and along the east-rift zonehave the same incompatible element ratios, a feature consistentwith a comagmatic origin. Competing hypotheses for the relationnshipbetween the picrites and basalts are accumulation of olivinein a basaltic melt similar in composition to the Halemaumaubasalts or fractionation of olivine from a primitive picriticmelt to produce the basalts. The results of thermodynamic modelingand major element trends on element ratio diagrams disprovethe first hypothesis but are consistent with the second, fractionation.Thermodynamic modeling provides further tests of the fractionationhypothesis in the form of predicted phase compositions. Predictedand observed phase compositions do not differ significantly.As predicted by modeling, late crystallizing augite and plagioclasein the east-rift picrites are similar to early crystallizingaugite and plagioclase in the Halemaumau basalts. These resultsare all consistent with the hypothesis that primitive picriticmelts with more than 15% MgO fractionated along a liquid lineof descent from the Hiiaka picrites to the Halemaumau basalts.The original magma entered the volcano, probably near its baseon the ocean floor. Part ascended and differentiated under Halemaumauand part erupted at Hiika, 5 km down the east-rift with littlechemical modification. More extensively crystallized picritesthat represent fractionation stages between the Hiiaka Craterpicrites and the Halemaumau basalts erupted another 4?5 and19?5 km down rift from Hiiaka, ending the August, 1968 eruption.     

Petrology of middle proterozoic diabases and picrites from Lake Nipigon,Canada

Mafic rocks at Lake Nipigon provide a record of rift-related continental basaltic magmatism during the Keweenawan event at 1109 Ma. The mafic rocks consist of an early, volumetrically minor suite of picritic intrusions varying in composition from olivine gabbro to peridotite and a later suite of tholeiitic diabase dikes, sheets and sills. The diabase occurs primarily as two 150 to 200 m thick sills with a textural stratigraphy indicating that the sills represent single cooling units. Compositional variation in the sills indicates that they crystallized from several magma pulses.The diabases are similar in chemistry to olivine tholeiite flood basalts of the adjacent Keweenawan rift, particularly with respect to low TiO₂, K₂O and P₂O₅. The picrites have higher TiO₂, K₂O and P₂O₅ than the diabases and are similar to, but more primitive than, high Fe-Ti basalts which erupted early in the Keweenawan volcanic sequence.All of the rocks crystallized from fractionated liquids. The picrites are cumulate rocks derived at shallow crustal depths from a magma controlled predominantly by olivine fractionation. Picritic chills are in equilibrium with olivine phenocrysts of composition Fo₈₀ and are interpreted to represent the least evolved liquids observed. The parental magma of the picrites was probably Fe rich relative to the parental magma of the diabase. The diabase sills crystallized from an evolved basaltic liquid controlled by cotectic crystallization of plagioclase and lesser olivine and pyroxene.The emplacement of dense olivine phyric picritic magmas early in the sequence, followed by later voluminous compositionally evolved magmas of lower density suggests the development of a crustal density filter effect as the igneous event reached a peak. Delamination of the crust-mantle interface may have resulted in the transition from olivine controlled primitive magma to fractionated magma through the development of crustal underplating.     

Alkali feldspar liquid equilibrium relationships in peralkaline oversaturated systems and volcanic rocks

A critical review of both experimental data and natural rock sequences allows a comparison of liquid lines of descent in peralkaline oversaturated systems. Available experimental data have been used to draw the liquidus surfaces and the fractionation curves in various planes including the alkali feldspar join. The thermal valley becomes steeper, the fractionation curves being channeled faster towards the thermal valley with increasing alkali-silica ratio. The slope of the thermal valley varies so that the higher the peralkalinity, the more sodic the most fractionated liquids. Rock series from Easter Island, Gran Canaria, Pantellaria Island, Boina (Afar), Fantale (Ethiopia) and Menengai (Kenya) in which there is good evidence for alkali feldspar fractionation, show that such variations do occur in nature. Factors other than the peralkalinity of the series, such as additional components and changes in the confining pressure are shown to affect the orientation and location of the thermal valley.     

The Peralkaline Volcanic Suite of Aden and Little Aden, South Arabia

The volcanic rocks of Aden, Little Aden, and Ras Imran, heredesignated as belonging to the Aden Volcanic Series, were eruptedthrough central-vent, strato-volcanoes about 5 m.y. ago. Inits major element chemistry the Aden Volcanic Series is intermediatebetween the alkaline and tholeiitic associations, and this isdemonstrated by comparing it with the alkaline suite of Hawaiiand the tholeiitic series of Thingmuli, Iceland. It is proposedthat the most acceptable ‘parental’ magma is a mildlyalkaline olivine basalt which, on fractionation, produced aseries ranging from trachybasalts through trachyandesites andtrachytes to rhyolites. These rhyolites are peralkaline as themolecular proportion of alumina is less than that of the combinedalkalis, and are comenditic as the series is poor in normativefemic constituents. Trace element data suggest that the peralkalinesilicic eruptives are chemically comparable with those of MayorIsland, New Zealand, where a mildly alkaline olivine basaltparent has also been postulated. Although the age of eruption of c. 5 m.y., given by K-Ar measurements,is entirely consistent with an age deduced from geomorphologicalcriteria, an ⁸⁷Sr/⁸⁶Sr versus ⁸⁷Rb/⁸⁶Sr isochron plot suggeststhat the series is related to a thermal event some 20-30 m.y.older than the age of eruption. As this earlier age correspondsdirectly to the age of the previous magmatic episode, the eruptionof the Yemen Trap Series, the upper part of which is petrologicallysimilar to the Aden Volcanic Series, and as the initial ⁸⁷Sr/⁸⁶Srratios suggest that the magma originated in the mantle, it isproposed that the most acceptable petrogenetic scheme, whichwould also explain the anomalously old Rb-Sr age, is: (a) Partialfusion in the upper mantle giving rise to the alkaline YemenTrap Series, (b) After the cessation of surface activity, alarge body of magma existed in the upper mantle and this magma,on crystallizing, fractionated to produce a layered sequence,(c) About 5 m.y. ago some event, either pressure relief or furtherthermal activity, resulted in the partial remelting of thisfractionated plutonic sequence and the liquids so formed reachedthe surface without significant mixing or chemical fractionation.     

Fractional crystallization and magma mixing in the Tigalak layered intrusion,the Nain anorthosite complex,Labrador

The Tigalak intrusion is a dominantly dioritic layered body, about 80 km² in area, which ranges in composition from norite to granodiorite. Local areas of the layered rocks display upward fractionation from norite to ferrodiorite. Periodic reversals of mineral composition trends record the emplacement of less fractionated dioritic magma. Heterogeneous mixtures of dioritic and granodioritic rocks occur widely in mappable lenses and layers that alternate up section and along the strike with more uniformly layered rocks. In these mixtures, chilled dioritic pillows occur abundantly in a hybrid cumulate matrix of granodiorite to diorite composition. Cross-cutting granodioritic dikes grade upward into stratigraphically-bound lensoid masses of the hybrid cumulates. It appears that the hybrid rocks formed as a result of the emplacement of the granodioritic magma through lower cumulates into the dioritic magma chamber and that the dioritic pillows represent chilled bodies of Ferich dioritic magma that commingled with cooler granodioritic magma and settled to the floor of the Tigalak magma chamber. The restricted distribution of these mixtures of hybrid cumulates and chilled pillows indicates that mixing between granodioritic and dioritic liquids was limited in time and lateral extent. Periodic injections of granodioritic liquids may have collected as a separate layer below the roof of the magma chamber and above dioritic magma.     

The behavior of noble gases in silicate liquids: Solution,diffusion, bubbles and surface effects,with applications to natural samples

Solubilities of noble gases in five natural silicate liquids as a function of temperature and partial pressure at 1 atm total pressure were determined and diffusion coefficients of the noble gases were measured in a tholeiite basalt at 1350°C. Solubilities of noble gases in silicate liquids obey Henry's law and are a strong function of composition and/or physical properties of the liquids. Solubility is greatest in less dense, more silica-rich liquids. Solubility is highest for the light gases and is related to the radius of the gas atom according to K_i = aexp (−br²_i). Temperature dependence of solubility is weak, but in general solubility increases with increasing temperature yielding positive enthalpies of solution. Diffusion coefficients in a basalt liquid at 1350°C show more or less the same linear relationship with r² as solubility and are larger than what would be expected from extrapolation of values determined at lower temperatures. A large percentage of samples of andesite composition had bubbles that contained gas which was fractionated from the gas of the experiment. Concentrations of noble gases in samples equilibrated on Pt wire loops correlate with the surface/volume ratio, suggesting that surfaces of silicate liquids can accommodate more noble gases than the liquid proper. Solubility fractionation is a valid process to account for certain patterns in marine basalts. The density of silicate liquids appears to be a good predictor of noble gas solubility in these liquids.     

Petrographic and Geochemical Evidence for Magma Mixing and Crustal Contamination in the Post-Collisional Calc-Alkaline Yozgat Volcanics,Central Anatolia,Turkey

Petrographic, major-oxide, and trace-element data are presented for the Yozgat volcanics. These rocks range in composition from basalts through basaltic andesites and andesites to dacites. Major-oxide variations are largely explicable in terms of fractional crystallization, involving removal of observed phenocrysts and microphenocrysts. However, complex zoning patterns and resorbtion phenomena shown by phenocrysts in these lavas, and observed epitaxitic pyroxene growth around quartz xenocrysts imply that they are hybrids formed by a mixing process. In addition, observed enrichments in crustal elements such as K, Rb, Ba, Sr, and P provide clear evidence for the crustal assimilation of granitoid and metasedimentary xenoliths. The following model is suggested for the evolution of the Yozgat volcanics. The primitive magma underwent fractionation in an intracrustal magma chamber to yield more evolved liquids. Influx of hot, primitive magma into the magma chamber promoted vigorous convection-crustal assimilation and eruption of the volcanic rocks in the study area.     

Petrology and evolution of transitional alkaline — sub alkaline lavas from Patmos,Dodecanesos, Greece: evidence for fractional crystallization,magma mixing and assimilation

Petrographic, mineral chemical and whole-rock major oxide data are presented for the lavas of the Main Volcanic Series of Patmos, Dodecanesos, Greece. These lavas were erupted about 7 m.y. ago and range in composition from ne-trachybasalts through hy-trachybasalts and trachyandesites to Q-trachytes. To some extent, the ne-trachybasalts are intermediate in composition to the alkaline lavas found on oceanic islands and the calc-alkaline lavas of destructive plate margins. Major oxide variation is largely explicable in terms of fractional crystallization involving removal of the observed phenocryst and microphenocryst phases viz. olivine, plagioclase, clinopyroxene and Ti-magnetite in the mafic lavas, plagioclase, clinopyroxene, mica and Ti-magnetite in the evolved lavas. Apatite, which occurs as an inclusion in other phenocrysts or as microphenocrysts must also have been removed. However, mass balance calculations indicate that the chemistry of the hy-trachybasalts is inconsistent with an origin via fractional crystallization alone and the complex zoning patterns and resorbtion phenomena shown by phenocrysts in these lavas show that they are hybrids formed by the mixing of 80-77% ne-trachybasalt with 20–23% trachyandesite. It is estimated that the mixing event preceded eruption by a period of 12 h-2 weeks suggesting that mixing triggered eruption. Combined fractionation and mixing cannot explain the relatively low MgO contents of the hy-trachybasalts and it is concluded that assimilation also occurred. Assimilation, and especially addition of volatiles to the magmas, may be responsible for the evolutionary trend from ne-normative to hy-normative magmas and was probably facilitated by intensified convection resulting from mixing. A model is presented whereby primitive magma undergoes fractionation in an intracrustal magma chamber to yield more evolved liquids. Influx of hot primitive magma into the base of the chamber facilitates assimilation, but eventually mixing yields the hy-trachybasalts and finally the ne-trachybasalts are erupted.     

A comparative study of olivine and clinopyroxene spinifex flows from Alexo,abitibi greenstone Belt,Ontario, Canada

A petrological and geochemical study of an olivine and of a clinipyroxene spinifex textured flow, from Alexo, indicates that the initial liquid in both flows probably came from the same mantle melting event and that the source was incompatible element depleted. The starting liquid of the clinopyroxene flow had experienced more olivine fractionation (10%) prior to its emplacement at Alexo, than the initial liquid of the olivine spinifex flow. The development of each of the textural and compositional zones in the flows can be modelled by means of crystal fractionation. In the case of the clinopyroxene flow the B-zone is formed by the fractionation of olivine, low-Ca pyroxene and chromite. An unusual feature of the Alexo clinopyroxene flow is presence of a peridotitic komatiite above the pyroxene cumulate layer, where a basaltic komatiite would usually be present. The presence of the peridotitic komatiite suggests an influx of new magma and hence a dynamic model for the flow. The composition of the clinopyroxene spinifex zone represents a mixture of clinopyroxene plus liquid, rather than simply a frozen liquid. This could happen if the clinopyroxene needles grew stalactitelike from the chilled upper surface of the flow into a flowing basaltic liquid. In the olivine spinifex flow the zones can be modelled as frozen liquids in the A₂-zone, as initial liquid which has fractionated 30% olivine in the A₃-zone and as liquid plus 50% olivine in the B-zone. But, if the clinopyroxene spinifex developed by stalactite growth of clinopyroxene needles into the a flowing liquid, the possibility that the olivine spinifex represent fractionated liquid plus stalactite olivines arises.     

Petrology of Submarine Lavas from Kilauea's Puna Ridge, Hawaii

We have studied 30 quenched tholeiitic lava flows recoveredby 20 dredge hauls and one submersible dive along Puna Ridge,the submarine part of the East Rift Zone of Kilauea Volcano,Hawaii Glass grains from numerous additional flows were recoveredin turbidite sands cored in the Hawaiian Trough. These quenchedlavas document variable primary magma compositions; olivineand multiphase crystallization and fractionation; degassing;wall-rock stoping and assimilation; mixing in the crustal reservoirand the rift zone; entrainment of olivine xenocrysts from ahot, ductile, olivine cumulate body; and disruption of gabbrowallrocks in the rift zone. Glass grains in turbidite sands contain up to 15•0wt% MgO,in contrast to < 7•0wt% MgO for the sampled glass rindson lavas. The most forsteritic olivine phenocryst (F0₉₀₇) isin equilibrium with primary Kilauea liquid containing an average16•5 wt% MgO, but ranging from 13•4 to 18•4%.Lavas and glass grains have more restricted P₂O₅/K₂O and TiO₂/K₂Othan glass inclusions in olivine, because more diverse liquidstrapped as glass inclusions are mixed and homogenized beforeeruption. Variable trace element compositions in glass grainsand whole rocks indicate that the primary liquids form by partialmelting of mantle sources retaining clinopyroxene and garnet. Orthopyroxene xenocrysts formed at moderate pressures provideevidence for a sub-crustal staging zone. Chromite and olivinecrystallize in the crustal magma reservoir as the liquid coolsfrom an average 1346C to 1170C. Low viscosities of the primaryliquids (04 Pas) facilitate olivine settling, and the crystallizedolivine forms an olivine cumulate body at the base of the reservoir.Olivine is deformed as the hot ductile dunite body flows downand away from the summit. This flow drives instability of theHilina landslide on Kilauea. Dikes intrude the dunite, and magmaflowing through the dikes disaggregates and entrains olivinexenocrysts in Puna Ridge magmas. Primary liquids pond at or near the base of Kilauea's crustalreservoir because they are denser than more fractionated liquidsthat occupy the upper parts of the reservoir. The sulfur andwater contents of glass rinds indicate that fractionated liquidsnear the top of the reservoir degas at low pressure, a processthat increases their density and causes them to sink to levelswhere they mix with resident undegassed, near-primary liquid.The fractionated liquids near the top of the magma reservoiracquire excess Cl, owing to assimilation of hydrothermally alteredroofrocks. Magma flowing into the rift zone encounters and mixes with low-temperature,multiphase-fractionated melt. The mixed magmas typically containrare orthopyroxene, plagioclase as sodic as andesine, olivineas fayalitic as F0₇₅ and Fe-rich augite derived from the fractionatedmagma. Magma flowing through dikes also dislodged fragmentsof gabbroic wallrocks that occur as xenoliths. The interrelations in the Kilauean submarine lavas between hostglass and glass inclusion compositions, volatile contents andmineral chemistry reveal an extraordinarily complex sequenceof petrogenetic processes and events that are difficult or impossibleto determine in subaerial Kilauea lavas because of crystallization,reequilibration and degassing during or after their eruption. KEY WORDS: submarine lavas; petrology; Kilauea; Hawaii; magma mixing ^*Corresponding authorPresent address: Rosentiel School of Marine and Atmospheric Science, Division of Marine Geology and Geophysics, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149-1098, USA     

Experimental constraints on depths of fractionation of mildly alkalic basalts and associated felsic rocks: Pantelleria,Strait of Sicily

Pantelleria, Italy, is a continental rift volcano consisting of alkalic basalt, trachyte, and pantellerite. At 1 atm along the FMQ buffer, the least-evolved basalt (Mg #= 58.5% norm ne) yields olivine on the liquidus at 1,180° C, followed by plagioclase, then by clinopyroxene, and by titanomagnetite and ilmenite at 1,075°. After 70% crystallization, the residual liquid at 1,025° is still basaltic and also contains apatite and possibly kaersutite. A less alkalic basalt shows the same order of phase appearance. Glass compositions define an Fe-enrichment trend and a density maximum for anhydrous liquids that coincides with a minimum in Mg#.During the initial stages of crystallization at 1 atm, liquids remain near the critical plane of silica-undersaturation until, at lower temperatures, Fe-Ti oxide precipitation drives the composition toward silica saturation. Thus the qtz-normative trachytes and pantellerites typically associated with mildly ne-normative basalts in continental rifts could be produced by low-pressure fractional crystallization or by shallow-level partial melting of alkali gabbro. At 8 kbar, clinopyroxene is the liquidus phase at 1,170° C, followed by both olivine and plagioclase at 1,135°. Because clinopyroxene dominates the crystallizing assemblage and plagioclase is more albitic than at 1 atm, liquids at 8 kbar are driven toward increasingly ne-normative compositions, suggesting that higher-pressure fractionation favors production of phonolitic derivatives.Natural basaltic samples at Pantelleria are aphyric or contain 1–10% phenocrysts of plag olcpx or ol>cpx, with groundmass Fe-Ti oxides and apatite. The lack of phenocrystic plagioclase in two of the lavas suggests that crystallization at slightly higher PH₂O may have destabilized plagioclase relative to the 1-atm results, but there is no preserved evidence for significant fractionation at mantle depths as clinopyroxene is the least abundant phenocryst phase in all samples and contains only small amounts of octahedral Al. The liquid line and phenocryst compositions match more closely the 1-atm experimental results than those at 8 kbar.Although major-element trends in natural liquids and crytals reflect low-pressure fractionation, minor- and trace-element concentrations preserve evidence of multiple parental liquids. Scatter in variation diagrams exceeds that attributable to crystal accumulation in these phenocryst-poor rocks, and the large range in concentrations of P and Ti at high MgO contents cannot be produced by polybaric fractionation nor by mixing with coexisting felsic magmas. Sr and O isotope ratios rule out significant interaction with crystalline upper crust, Mesozoic shelf sediments, or Tertiary evaporites. Positive correlations of compatible and incompatible elements suggest that the basalts are not simply related to one another by closed-system fractional crystallization of a single parental magma. Increasing Ce/Yb with Ce suggests that these relations are not a product of mixing within a replenished magma chamber, nor of mixing with more felsic members of the suite, which have smaller Ce/Yb ratios. Low-pressure fractional crystallization of ol+cpx+ plag±oxides from slightly different parental magmas produced by varying degrees of melting of garnet-bearing peridotite is a possible scenario.Small and infrequently replenished magma reservoirs in this continental rift environment may account for the strongly differentiated nature of the Pantellerian basalts. There is no correlation between Mg# and eruptive frequency, in part because concentration of volatiles in residual liquids offsets the effect of Fe-enrichment on melt density, such that strong Fe-enrichment is no hindrance to eruption.     

Kinetic isotopic fractionation during diffusion of ionic species in water

Experiments specifically designed to measure the ratio of the diffusivities of ions dissolved in water were used to determine . The measured ratio of the diffusion coefficients for Li and K in water (D_Li/D_K = 0.6) is in good agreement with published data, providing evidence that the experimental design being used resolves the relative mobility of ions with adequate precision to also be used for determining the fractionation of isotopes by diffusion in water. In the case of Li, we found measurable isotopic fractionation associated with the diffusion of dissolved LiCl (D_⁷Li/D_⁶Li=0.99772±0.00026). This difference in the diffusion coefficient of ⁷Li compared to ⁶Li is significantly less than that reported in an earlier study, a difference we attribute to the fact that in the earlier study Li diffused through a membrane separating the water reservoirs. Our experiments involving Mg diffusing in water found no measurable isotopic fractionation (D_²⁵Mg/D_²⁴Mg=1.00003±0.00006). Cl isotopes were fractionated during diffusion in water (D_³⁷Cl/D_³⁵Cl=0.99857±0.00080) whether or not the co-diffuser (Li or Mg) was isotopically fractionated. The isotopic fractionation associated with the diffusion of ions in water is much smaller than values we found previously for the isotopic fractionation of Li and Ca isotopes by diffusion in molten silicate liquids. A major distinction between water and silicate liquids is that water surrounds dissolved ions with hydration shells, which very likely play an important but still poorly understood role in limiting the isotopic fractionation associated with diffusion.     

The differentiation of the Skaergaard Intrusion

Previous interpretations of the Skaergaard Intrusion suggested that differentiation involved extreme iron-enrichment but no silica-enrichment until a very late stage. This model is difficult to reconcile with petrological and geochemical evidence, with the behaviour of tholeiitic volcanic suites and with phase equilibria. We propose that the Skaergaard magma evolved on a trend of pronounced silica-enrichment after cumulus magnetite appeared at the top of the Lower Zone. At that stage, the magma was of ferrobasaltic composition with close to 50% SiO₂. The Middle and Upper Zones of the intrusion dominantly represent crystal accumulation during differentiation from ferrobasalt through iron-rich basaltic andesite and icelandite to rhyolite, a fractionation sequence common in tholeiitic volcanic provinces. This interpretation requires re-appraisal of the physical processes responsible for the differentiation. In particular, residual liquids became lower in density with fractionation and would have caused the Skaergaard magma chamber to have become compositionally zoned.     

The Mineralogy and Petrology of Mount Suswa, Kenya

Mount Suswa, a Quaternary volcano in the Rift Valley of Kenya,is composed of sodalite-trachytes, sodalite-phonolites, andphonolites, the majority of which are mildly peralkaline. Thelavas are predominantly feldspathic with phenocrysts of alkalifeldspar and, less commonly, slightly sodic augite and fayaliticolivine. The groundmass typically contains alkali feldspar,augite, titanomagnetite, and may in addition contain sodalite,nepheline, alkali amphibole, aenigmatite, and glass. The lavasof four stratigraphically distinct episodes can be distinguishedon the basis of mineralogy and chemical composition. These lavasare the products of at least three parental magmas, none ofwhich appears to be a derivative of the other. Each ‘magmatype’ represents an independent episode of magma generation,emplacement, and eruption. The order of eruption in the finalepisode corresponds to increasing peralkalinity and undersaturationwith respect to silica, and indicates that these lavas weregenerated via the tapping of a differentiating magma, with thefirst lavas being the least differentiated. Utilizing coexistingfeldspar, residual glass, and bulk rock compositions, the derivationof peralkaline phonolitic residual liquids from a trachyticparent is shown to be a process controlled by feldspar fractionation.     

A Model for Flood Basalt Vulcanism

The question of whether basaltic rocks in continental floodbasalt provinces are primary magmas or whether they are descendedin general from picritic parent magmas is reviewed. It is suggestedthat the latter is more likely to be correct on the evidenceof phase relations and the relative rareness of mantle materialswith appropriate Fe/Mg ratios. Major element variations in the residual liquids of fractionaland equilibrium crystallization of basaltic magmas are modelledfor a variety of crystallizing assemblages. It is concludedthat crystallization of olivine, clinopyroxene, and plagioclasehas a marked effect on buffering chemical change in many importantelements. It is this effect which accounts for the apparentuniformity of large volumes of flood basalts, not, as has sometimesbeen supposed, a series of implausible coincidences in the amountof material fractionated from each magma batch. It is furtherargued that much of the variation seen in basalts may be imposedby polybaric fractionation operating throughout crustal depths,that is at pressures up to at least 12 kb. Parental picriticmagmas rising from the mantle reach the surface in exceptionalareas of crustal thinning. More usually, however, it is suggestedthat they intrude the base of the crust as a series of sillswhich differentiate into upper gabbroic and lower ultramaficportions. Much of the ‘low pressure’ fractionationof basaltic magmas may take place in this deep crustal sillcomplex and evolved liquids are transmitted to the surface astheir density becomes sufficiently low. This implies that inareas of flood vulcanism a potentially large new contributionto the crust is made by under-plating, the volumes of concealedcumulates being at least as large as the amount of erupted surfacelava.     

Variation in rhyolitic magma composition linked with fractionation from a common source: insights from the Rotoiti eruption,Taupo Volcanic Zone,New Zealand

ABSTRACT

The Taupo Volcanic Zone (TVZ), New Zealand, is a well-documented volcanic arc characterized by explosive rhyolitic magmas within a series of caldera complexes that include the Okataina Volcanic Centre (OVC). New quartz melt inclusion and volcanic glass data from the 45 ka caldera-forming Rotoiti eruption within the OVC are compared to published studies. The new data are characterized by low K₂O (~1.5–3.5 wt.%), Rb (~30–70 ppm), Sr (~40–90 ppm), U (~0.5–2.5 ppm), and Ba (~300–1000 ppm) ranges that differ significantly from other OVC systems (~3.0–4.5 wt.% K₂O, ~80–150 ppm Rb, and ~2.5–5.0 ppm U). Most interestingly, the Rotoiti melt inclusion data measured in this study show a decrease in Rb, Sr, and U, although the fractionation trends originate from the same source point as published OVC data. This progressive decreasing trend is interpreted as an interaction with a less enriched rhyolitic melt (represented by the low Rb, Sr, and U of glasses) during fractionation processes from a common TVZ source. The established model for TVZ rhyolites is that they are extracted from a middle or upper crustal source (‘mush’ zone) prior to eruption. Adding to this model, new melt inclusion data suggest that all TVZ rhyolites are fractionated from this common TVZ source and, prior to eruption, the Rotoiti system was rejuvenated by this source (evidenced by the low REE glasses). Exactly what triggers the common TVZ source to fractionate remains unclear, but a proposed mechanism to account for this involves the successive melting of the upper crust by upwelling mantle induced by incremental subduction.