Morphology and emplacement of flows from the Deccan Volcanic Province,India

The present study is probably the first of its kind in the Deccan Volcanic Province (DVP) that deals in detail with the morphology and emplacement of the Deccan Trap flows, and employs modern terminology and concepts of flow emplacement. We describe in detail the two major types of flows that occur in this province. Compound pahoehoe flows, similar to those in Hawaii and the Columbia River Basalts (CRB) constitute the older stratigraphic Formations. These are thick flows, displaying the entire range of pahoehoe morphology including inflated sheets, hummocky flows, and tumuli. In general, they show the same three-part structure associated with pahoehoe flows from other provinces. However, in contrast to the CRB, pahoehoe lobes in the DVP are smaller, and hummocky flows are quite common. 'Simple' flows occur in the younger Formations and form extensive sheets capped by highly vesicular, weathered crusts, or flow-top breccias. These flows have few analogues in other provinces. Although considered to be a'a flows by previous workers, the present study clearly reveals that the simple flows differ considerably from typical a'a flows, especially those of the proximal variety. This is very significant in the context of models of flood basalt emplacement. At the same time, they do not display direct evidence of endogenous growth. Understanding the emplacement of these flows will go a long way in determining whether all extensive flows are indeed inflated flows, as has recently been postulated.Most of the studies relating to the emplacement of Continental Flood Basalt (CFB) lavas have relied on observations of flows from the CRB. Much of the current controversy surrounding the emplacement of CFB flows centers around the comparison of Hawaiian lava flows to those from the CRB. We demonstrate that the DVP displays a variety of lava features that are similar to those from the CRB as well as those from Hawaii. This suggests that there may have been more than one mechanism or style for the emplacement of CFB flows. These need to be taken into account before arriving at any general model for flood basalt emplacement.Editorial responsibility: T. Druitt     

Structure and emplacement of the Nandurbar–Dhule mafic dyke swarm,Deccan Traps,and the tectonomagmatic evolution of flood basalts

Flood basalts, such as the Deccan Traps of India, represent huge, typically fissure-fed volcanic provinces. We discuss the structural attributes and emplacement mechanics of a large, linear, tholeiitic dyke swarm exposed in the Nandurbar–Dhule area of the Deccan province. The swarm contains 210 dykes of dolerite and basalt >1 km in length, exposed over an area of 14,500 km². The dykes intrude an exclusively basaltic lava pile, largely composed of highly weathered and zeolitized compound pahoehoe flows. The dykes range in length from <1 km to 79 km, and in thickness from 3 to 62 m. Almost all dykes are vertical, with the others nearly so. They show a strong preferred orientation, with a mean strike of N88°. Because they are not emplaced along faults or fractures, they indicate the regional minimum horizontal compressive stress (σ ₃) to have been aligned ~N–S during swarm emplacement. The dykes have a negative power law length distribution but an irregular thickness distribution; the latter is uncommon among the other dyke swarms described worldwide. Dyke length is not correlated with dyke width. Using the aspect ratios (length/thickness) of several dykes, we calculate magmatic overpressures required for dyke emplacement, and depths to source magma chambers that are consistent with results of previous petrological and gravity modelling. The anomalously high source depths calculated for a few dykes may be an artifact of underestimated aspect ratios due to incomplete along-strike exposure. However, thermal erosion is a mechanism that can also explain this. Whereas several of the Nandurbar–Dhule dykes may be vertically injected dykes from shallow magma chambers, others, particularly the long ones, must have been formed by lateral injection from such chambers. The larger dykes could well have fed substantial (≥1,000 km³) and quickly emplaced (a few years) flood basalt lava flows. This work highlights some interesting and significant similarities, and contrasts, between the Nandurbar–Dhule dyke swarm and regional tholeiitic dyke swarms in Iceland, Sudan, and elsewhere. Editorial responsibility: J. White     

Palaeomagnetic investigations on the Deccan Traps from Chincholi,Mysore State,India

Summary 38 oriented samples of Deccan Traps have been collected from the neighbourhood of Chincholi, Mysore State, India. The Natural Remanent Magnetisation of these rocks has been studied using an astatic magnetometer. It has been found that these rocks are magnetically reversed, the mean magnetic direction being N154°E in declination and 61° down in inclination. Thermoremanance studies conducted on four specimens showed that two specimens with weak NRM and a high secondary magnetisation have Curie temperatures around 560°C for the NRM and exhibited partial reversal of TRM at room temperature, while two specimens with high NRM and with little secondary magnetisation have Curie temperatures much lower than 560°C for the NRM.     

Geological and geophysical investigations on Deccan Traps

Volcanic rocks occupy considerable regions in the western portion of India, attaining a maximum thickness of 7000′ near Igatpuri. These rocks are essentially basaltic in nature and are generally referred to as plateau basalts. An attempt has been made in this paper to present some results of geological and geophysical investigations carried out in the Deccan Traps. Three areas (Ajanta - Long. 75″41′ -75° 45′ E, Lat. 20° 32′ - 20° 35′ 15″ N, 18 sq. miles in area; Ellora -Long. 75″ 11′ - 75° 16′ E, Lat. 20° 1′ - 20° 9′ N, 80 sq. miles in area; and Chincholi - Long. 77° 22′ - 77° 30′ E, Lat. 17° 22′ -17° 30′ N, 50 sq. miles in area) have been chosen for this study because of their geological setting. A large number of field specimens have been collected for petrographic study. This is supplemented by examination of microsections and chemical analyses of a few traps. In the Chincholi area where the trap overlies the granites, limestones seem to intervene in between trap and granites. With a view to estimate the possible thickness of the limestone beds, the distribution of intensity of magnetic field in a portion of the area has been studied with a magnetometer. Magnetic susceptibilities in case of few specimens have also been studied. Elastic constants of Deccan Traps have been determined for fifty specimens, employing the Wedge Method. These are further correlated with textural features and porosity values. Such an integrated geological and geophysical investigation on Deccan Traps is bound to reveal some interesting results.     

玄武岩内非均质性对地震资料的影响:印度德干圈闭的一个研究实例

Hydrocarbon exploration interests have renewed the need for developing new sub basalt imaging techniques. One of the most important problems encountered today is seismic imaging below basalt. In recent years, this problem appears to have been overcome partly by using long offset seismic data. However near offset data are yet to be fully utilised due to the complex waveform caused by the surface as well as internal heterogeneity of the basalts. The near normal incidence data, which influence the sub-basalt imaging, are highly useful to understand the internal structure within a basalt layer. The use of converted waves for such targets has been proposed as an alternative in a rather homogeneous basalt layer. With a few synthetic modelling exercises here we highlight the practical difficulties in dealing with more realistic and heterogeneous basalt flow. Full waveform seismograms are computed to understand the effects of intra-trappean sediments on the seismic data. A case study from the Deccan Traps of India is presented in this paper. First, we discuss the effects of intercalated sediments on the overall seismic image. Later, the sonic log data from the field are used to compute the full wave-field response using the reflectivity method and compared with the field data. The feasibility of using mode converted waves （P to S and vice-versa at the top and bottom basalt interfaces） for sub-basalt imaging in Kutch region is discussed through a series of velocity-depth profiles. By comparing with the field data we demonstrate that the effects of multiple thin layering within the basalt can strongly deteriorate the image we seek to interpret and exploit.     

玄武岩内非均质性对地震资料的影响:印度德干圈闭的一个研究实例(英文)

油气勘探需要重新开发玄武岩下目的层的成像技术。我们今天遇到的最重要的问题之一是玄武岩下地震成像。近年来,由于利用长炮检距这个问题部分已经得到解决。然而,由于地表以及玄武岩的内部非均质性引起的复杂波形,近炮检距的资料仍然不能充分地被利用。影响玄武岩下成像的近垂直入射资料对于了解玄武岩层内部结构是十分有用的。对比较均匀的玄武岩目标层可选用转换波。这里我们利用几个合成模型例子重点描述了更接近现实的非均质玄武岩流引起的实际困难。模拟计算了全波地震记录以有助于了解玄武岩内沉积物对地震资料的影响。本文介绍了印度德干圈闭的一个研究实例。首先探讨了夹层沉积物对整个地震成像的影响。其次利用该区声测井资料以反射系数法计算全波场响应与实际地震资料对比, 通过一系列速度-深度剖面探讨了利用模式转换波(顶底玄武岩界面上的P波转化为S波或相反)对印度库奇地区玄武岩下成像的可行性。通过与野外资料相比较证明玄武岩中多个薄层的效应严重恶化我们所要解释和利用的图象的质量。     

Palaeogeography,geomorphological setting and groundwater possibilities in the Deccan Traps of Western Maharashtra

Deccan Traps are the most extensive geological formations of Deccan Peninsula with the exception of only the metamorphic and igneous complex of Archaean age. Based on their mode of emplacement, geomorphic setting and hydrogeological behaviour over an area of about 5,000 sq. km the authors have classified the Deccan Traps of western Maharashtra into 3 groups, namely, (1) The Deccan Traps of Dhulia district, characterised by numerous dolerite dykes, (2) Areally extensive trap flows of Sholapur and Osmanabad districts resulting from slow and quiescent type of flood eruption occupyng the gently undulating terrain, and (3) the traps of Kolaba, Thana and Bombay-Poona regions characterised by intertrappean sediments, dolerite dykes and volcanic ash beds, indicative of violent outbursts resulting in the Sahyadri geomorphologic unit. The groundwater possibilities in the three groups are to a great extent governed by the nature and constitution of the individual flows. The massive traps with their fracture porosities, the vesicular traps with their minutely interconnected and partly filled vesicles and the intertrappen sediments with their primary porosities play a decisive role in determining the groundwater possibilities in them. In Dhulia district the dolerite dykes to a great extent control the movement of groundwater, and success or otherwise of the well field area depends very much upon its location with reference to adjacent dykes. Areally extensive thick vesicular traps with their gentle dips towards east, in Sholapur district, have to be explored for possible artesian conditions in the downdip directions of the trappean units to be tapped. In the case of Poona, Thana and Kolaba districts, exploratory drilling based on geophysical data (to delineate the nature and extent of water bearing horizons) has to be resorted to. It is, therefore, imperative to sub-divide at this stage Taylor’s Single Unit of Deccan Trap Groundwater Province into 3 Sub-Provinces, based on geomorphological, geological and geohydrological setting in the region of western Maharashtra of the present investigation.     

Study of the Deccan Traps of Cambay Basin by geophysical methods

In the attempt to study the buried Deccan Trap layers in the Cambay Basin, the ground magnetic surveys have not been very useful as the data combine the effect due to the crystalline basement and the Trap thickness. In some parts of the basin, some reflections in the seismograms obtained in the course of seismic surveys, could be correlated to the Trap surface. These can be tied with wells drilled in the basin upto the Traps. The synthesis of the gravity and seismic data has enabled us to prepare a map of the Trap surface in the Cambay basin. The depth of the Trap surlace increases from about 2000 m in the northern part of the basin to about 600 m in its deepest part near Broach. The Trap surface rises gradually south of Narbada in an average direction of SE with depths running from 2500 m to 500 m. The interpretation of the gravity anomalies, assuming their cause to be the variations in the thickness of the Trap, has enabled the determination of the average thickness of the Traps in the basin. The maximum thickness of the Trap is in the central part of the basin and is estimated to be about 2.4 km. The Traps appear to gradually taper towards the flanks of the basin.     

Morphology of rubbly pahoehoe (simple) flows from the Deccan Volcanic Province: Implications for style of emplacement

Lava flows with preserved bases and brecciated upper crusts constitute a morphological type that differs in character from typical pahoehoe and a'a: such flows have been reported from many provinces around the world. Previous studies had referred to these flows informally as ‘pahoehoe flows with rubbly tops’, ‘broken-top pahoehoe’ and ‘rubbly pahoehoe’. Recent studies have formally applied the latter term to describe parts of the well-studied Laki flow in Iceland as well as flows from the Columbia River Basalt province. Rubbly pahoehoe flows are abundant in the upper stratigraphic formations of the Deccan Volcanic Province (DVP), and are more commonly known as simple flows. This study presents detailed observations of such flows from various parts of the DVP and discusses their implications for understanding flow emplacement. These flows, which appear to be single units at the outcrop-scale, are generally much thicker and significantly more extensive than individual pahoehoe lobes that dominate the lower formations of the Deccan stratigraphy. They are characterised by preserved, gently undulating tachylitic bases but variably disrupted crustal zones that grade into flow-top breccias. The breccias are constituted of highly vesicular and oxidised fragments of varying sizes that appear to have been derived from previously formed pahoehoe crusts. Previous work has indicated that the morphology of these flows might be related to initial inflation, accompanied by rapid volatile exsolution and an increase in effusion rate and/or viscosity with time. This agrees reasonably well with the qualitative and quantitative models of emplacement developed for the Laki flow. The abundance of such flows in the upper formations of the Deccan stratigraphy clearly hints at a significant shift in the nature of the Deccan eruptions; this could be indicative of higher eruption rates during this period. This, in turn, raises the possibility of hazardous impact on the climate during the eruption of these flows, which is also discussed in the paper.     

Mafic alkalic magmatism in central Kachchh,India: a monogenetic volcanic field in the northwestern Deccan Traps

Magmatism in Kachchh, in the northwestern Deccan continental flood basalt province, is represented not only by typical tholeiitic flows and dikes, but also plug-like bodies, in Mesozoic sandstone, of alkali basalt, basanite, melanephelinite and nephelinite, containing mantle nodules. They form the base of the local Deccan stratigraphy and their volcanological context was poorly understood. Based on new and published field, petrographic and geochemical data, we identify this suite as an eroded monogenetic volcanic field. The plugs are shallow-level intrusions (necks, sills, dikes, sheets, laccoliths); one of them is known to have fed a lava flow. We have found local peperites reflecting mingling between magmas and soft sediment, and the remains of a pyroclastic vent composed of non-bedded lapilli tuff breccia, injected by mafic alkalic dikes. The lapilli tuff matrix contains basaltic fragments, glass shards, and detrital quartz and microcline, with secondary zeolites, and there are abundant lithic blocks of mafic alkalic rocks. We interpret this deposit as a maar-diatreme, formed due to phreatomagmatic explosions and associated wall rock fragmentation and collapse. This is one of few known hydrovolcanic vents in the Deccan Traps. The central Kachchh monogenetic volcanic field has >30 individual structures exposed over an area of ∼1,800 km² and possibly many more if compositionally identical igneous intrusions in northern Kachchh are proven by future dating work to be contemporaneous. The central Kachchh monogenetic volcanic field implies low-degree mantle melting and limited, periodic magma supply. Regional directed extension was absent or at best insignificant during its formation, in contrast to the contemporaneous significant directed extension and vigorous mantle melting under the main area of the Deccan flood basalts. The central Kachchh field demonstrates regional-scale volcanological, compositional, and tectonic variability within flood basalt provinces, and adds the Deccan Traps to the list of such provinces containing monogenetic- and/or hydrovolcanism, namely the Karoo-Ferrar and Emeishan flood basalts, and plateau basalts in Saudi Arabia, Libya, and Patagonia.     

Cretaceous volcanic rocks of the South Tethyan suture zone, Pakistan: implications for the Réunion hotspot and Deccan Traps

The location, ages, and geochemical characteristics of marine volcanic rocks preserved in the South Tethyan suture zone of Pakistan suggest that the Réunion hotspot was active off northwestern Greater India well before the emplacement, far to the south, of the Deccan flood basalts, the great bulk of which were erupted at 65-66 Ma and are widely believed to be associated with the hotspot’s plume-head phase. Most of the suture zone samples have Nd-Pb-Sr isotopic ratios (e.g. age-corrected ?_Nd(t)=+3.0 to +4.6) close to those expected for modern-type Réunion source mantle in the Late Cretaceous, and their incompatible element patterns resemble those of recent Réunion shield lavas. ⁴⁰Ar-³⁹Ar incremental heating yields ages of 73.4-72.0 Ma. Nevertheless, unless even older ages are discovered among the suture zone rocks, a pre-Deccan marine phase of Réunion hotspot activity on the Tethyan side of Greater India can be accommodated within the framework of the plume-head model.     

Geology of the saucer-shaped sill near Mahad, western Deccan Traps, India, and its significance to the Flood Basalt Model

An ～22-m-thick saucer-shaped sill occurs near Mahad and is exposed as a curvilinear, miniature ridge within the Deccan Traps. The sill has variable dips (42–55°). It has a 7.1-km long axis and 5.3 km short axis (aspect ratio of 1.4) and is larger than the MV sill of the Golden Valley sill complex, South Africa and the Panton sill, Australia. The sill has distinct glassy upper and lower chilled margins with a coarse-grained highly jointed core. The samples from the margin are invariably fractured and iron stained because of deuteric alteration. The rock from the sill is plagioclase-phyric basalt. At least three thick sill-like apophyses emanate from the base of the main sill. The apophyses change direction because of bending and thinning from a horizontal concordant sheet at the top to a discordant inclined form that bends again to pass into a lower horizontal concordant sheet. We interpret such features as ‘nascent saucer-shaped sills’ that did not inflate to form nested sills. Geochemically, the sill consists of poorly differentiated tholeiitic basalt that has a restricted geochemical range. Critical trace element ratios and primitive mantle normalised trace and REE patterns indicate that the sills have geochemical affinities to the Poladpur chemical type and that the pahoehoe flow they intrude belongs to the Bushe Formation. Calculated magmatic overpressures during sill emplacement range from 8.4 to 11.3 MPa (for Young’s modulus E?=?5 GPa) and 16.7 to 22.5 MPa (for E=10 GPa) and depth to magma chamber ranges from 8.5 to 11.5 km (E?=?5 GPa) and 17.1 to 22.9 km (E?=?10 GPa), consistent with petrological and gravity modelling. The volume of the Mahad sill is approximately 276 km³ and is constant irrespective of the variations in the values of host-rock Young’s modulus. In 1980, Cox (J Petrol 21:629–650, 1980) proposed a conceptual model of the crust–mantle section beneath the Karoo CFB which is considered as the fundamental model for flood basalt volcanism. Our paper confirms the presence of a sill plus the inferred substructure beneath Mahad that are compatible with predictions of that model. In LIPS, saucer-shaped sills are formed in areas experiencing extensional tectonics where processes such as the Cook–Gordon delamination and Dundurs elastic extensional mismatch between layered sedimentary rocks or lava flows are responsible for the deflection of dykes into sills. A similar process is envisaged for the formation of the Mahad sill.     

40Ar/39Ar dating on volcanic rocks of the Deccan Traps,India

⁴⁰Ar/³⁹Ar dating results on seven volcanic rocks from four areas of the Deccan Traps, India, suggest that volcanic activity more than 70 Ma ago might have occurred at least in limited areas.In the Igat Puri area, the uppermost flow shows an⁴⁰Ar/³⁹Ar age of 63 Ma, whereas a lower flow has an age of around 82–84 Ma.⁴⁰Ar/³⁹Ar ages of samples from the Bombay area also seem to favor the occurrence of volcanic activity more than 70 Ma ago. One rhyolite dyke from the Osam Hill in the Girnar Hill area shows a well-defined plateau age of 68 Ma, whereas two tholeiitic basalts from the Mahabaleshwar area indicate a total⁴⁰Ar/³⁹Ar age of around 63–64 Ma, though they show the effect of secondary disturbance in the age spectra.The volcanic activity(ies) more than 70 Ma ago may correspond to precursory one(s) for the main volcanic activity around 65 Ma ago in the Deccan Traps.     

Relationships between crustal contamination and crystallisation in continental flood basalt magmas with special reference to the Deccan Traps of the Western Ghats, India

The upper part of the Deccan Traps sequence (Bushe to Mahabaleshwar Formations) shows a statistically significant tendency for the most mafic lavas to be the most contaminated by crustal materials. This is the reverse of the relationship shown by suites evolving by contamination accompanied by fractional crystallisation (AFC). The observed correlations (e.g. between Mg-number and Sr isotope initial ratios) are partly an accidental consequence of the fact that the most mafic lavas are more abundant in the lower part of the sequence, while contaminant availability declines in the upper part. It is probable, however, that the correlations are augmented by increased contamination of hotter magma batches during ascent through dykes, a process during which fractional crystallisation is suppressed by magmatic turbulence. The absence of AFC relationships suggests that most of the contamination took place during the ascent stage rather than in a magma chamber. Other continental flood basalt provinces such as the Parana and Etendeka do show AFC relationships, and it is speculated that this may be a result of magma chamber contamination coupled with flow rates which prevent contamination during ascent.     

Tumuli and associated features from the western Deccan Volcanic Province, India

Whale-back-shaped uplifts called "tumuli" are common in the pahoehoe flows of the western Deccan Volcanic Province (DVP). Although they usually occur in hummocky flows, they are also associated with thicker sheet lobes. They have been subjected to a detailed morphometric and petrographic study for the first time. The tumuli are characterised by positive relief and "lava-inflation clefts" occupied by squeeze-ups. They display elongate as well as equant forms; some are constituted of a single flow lobe, whereas others display multiple flow lobes. Some tumuli appear to have developed along anastomosing tube systems. The detailed study of one of the tumuli reveals considerable petrographic and textural variations among the constituent flow units. Some of these, such as the enrichment of phenocrysts in squeeze-ups and breakouts, could be related to the emplacement dynamics of the tumulus. All the observed tumuli display much evidence of inflation or endogenous growth. Field observations and measurements reveal that the tumuli and associated pahoehoe features display a close similarity with their Hawaiian counterparts. This is a very significant observation since it points out to a similarity in nature and style of eruptions in Hawaii and at least in the western part of the DVP. This has an important bearing on determining the short, medium and long-term effusion rates in the Deccan; however, any concrete inference will have to await systematic volcanological studies of the lava features in the DVP.     

A reappraisal of ignimbrite emplacement: progressive aggradation and changes from particulate to non-particulate flow during emplacement of high-grade ignimbrite

We propose a mechanism by which massive ignimbrite and layered ignimbrite sequences — the latter liable to have been previously interpreted as multiple flow units-form by progressive aggradation during sustained passage of a single particulate flow. In the case of high-temperature eruptive products the mechanism simplifies interpretation of problematic deposits that exhibit pronounced vertical and lateral variations in texture, including between non-welded, eutaxitic, rheomorphic (lineated) and lava-like. Agglutination can occur within the basal part of a hot density-stratified flow. During initial incursion of the flow, agglutinate chills and freezes against the ground. During sustained passage of the flow, agglutination continues so that the non-particulate (agglutinate) layer thickens (aggrades) and becomes mobile, susceptible to both gravity-induced motion and traction-shear imparted by the overriding particulate part of the flow. The particulate to non-particulate (P-NP) transition occurs in and just beneath a depositional boundary layer, where disruptive collisions of hot viscous droplets give way, via sticky grain interactions, to fluidal behavior following adhesion. Because they have different rheologies, the particulate and non-particulate flow components travel at different velocities and respond to topography in different ways. This may cause detachment and formation of two independent flows. The P-NP transition is controlled by factors that influence the rheological properties of individual erupted particles (strain rate, temperature, and composition including volatiles), by cooling and volatile exsolution during transport, and by the particle-size population and concentration characteristics of the depositional boundary layer. At any one location along the flow path one or more of these can change through time (unsteady flow). Thus the P-NP transition can develop momentarily or repeatedly during the passage of an unsteady flow, or it can occur continuously during the passage of a quasi-steady flow supplied by a sustained explosive eruption. Vertical facies successions developed in the deposit (high-grade ignimbrite) reflect temporal changes in flow steadiness and in material supplied at source. The P-NP transition is also influenced by factors that affect flow behaviour, such as topography. It may occur at any location laterally between a proximal site of deflation (e.g. a fountain-fed lava) and a flow's distal limit, but it most commonly occurs throughout a considerable length of the flow path. Up-sequence variations in welding-deformation fabric (between oblate uniaxial to triaxial and prolate) reflect evolving characteristics of the depositional boundary layer (e.g. fluctuations from direct suspension-sedimentation to deposition via traction carpets or traction plugs), as well as possible modifications resulting from subsequent, post-depositional hot loading and slumping. Similar processes can also account for lateral lithofacies gradations in conduits and vents filled with welded tuff. Our consideration of high-grade ignimbrites has implications for ignimbrite emplacement in general, and draws attention to the limitations of the widely accepted models of emplacement involving mainly high-concentration non-turbulent transport and en masse freezing of high-yield-strength plug flows.     

Origin of contemporaneous tholeiitic and K-rich alkalic lavas: a case study from the northern Deccan Plateau,India

Concurrently erupted, alternating Deccan Trap flows of tholeiitic and potassic alkalic basalt outcrop along both banks of the Narmada River near Navgam. Nearby, around Rajpipla, early tholeiites are overlain by K-rich alkalic flows and intrusives, which are themselves cut by later tholeiitic dikes. Nd and Sr isotopic ratios of a wide variety of rocks from both areas form a single correlated array, which reflects mixing between positive ε_JUV and negative ε_JUV endmembers. There is an almost complete overlap between values for tholeiitic and alkalic samples; thus, both alkalic and tholeiitic primary magmas must have been produced that were isotopically much alike. A Rajpipla rhyolite also falls on the array, near the midpoint. For positive values of ε_JUV(T) the array is indistinguishable from that defined by the lower group of tholeiites at Mahabaleshwar, some 450 km to the south, implying a similar or identical high ε_JUV mantle source. The negative ε_JUV component, apparently different from either of the two at Mahabaleshwar, may have been continental crust or enriched mantle. Both alkalic and tholeiitic groups display wide overlapping ranges in incompatible elements other than K, Rb, and Ba—particularly in Sr and Nb. This partial decoupling of incompatible elements, in conjunction with the isotopic similarity between the two classes of rocks, is strongly suggestive of an enrichment event in portions of the mantle source shortly before magmatism.     

Nomenclature and petrochemistry of the peralkaline oversaturated extrusive rocks

Four-hundred and twenty-one analyses of quartz-normative, peralkaline, extrusive rocks have been collected from the literature and from unpublished sources and are used to examine chemical variation in this group of rocks. Comparisons are particularly made between the full body of data and the variations recorded in the non-hydrated obsidians alone byMacdonald andBailey (1973). It is argued that the compositions of the magmas which formed these obsidians and those which subsequently crystallised were similar as regards the major oxides SiO₂, Al₂O₃, FeO + Fe₂O₃, Na₂O and K₂O. Marked variations in the abundances of the minor oxides CaO and TiO₂ are shown to be a result of geographical location. Small but significant differences in the distribution of Al and Fe as a function of normative quartz can be recognised between various pantelleritic suites. A new classificatory scheme is proposed, based on the iron (as FeO) and Al₂O₃ contents. This is simpler than previously employed normative classifications, is more applicable to crystalline rocks, and, happily, in 95 % of cases gives the same rock name as the normative system.