New evidence for two sharp replacement fronts during albitization of granitoids from northern Aravalli orogen, northwest India

We present new evidence of infiltration metasomatism in granitoids that were albitized in a process that produced two sharp replacement fronts, both of which are clearly visible in the field. The two fronts advanced through the original granite simultaneously, but at different rates. Here we focus mainly on the Ajitgarh intrusive in the northern Aravalli orogen of northwest India. This intrusion shows geographically well-defined metasomatic zones on the outcrop scale as well as a large volume of original ferroan granite, both of which were poorly preserved in most of the previously studied Khetri granites. Stage I metasomatism transformed the grey original granite to pink microcline–albite granite, and stage II converted the microcline–albite granite to white albite granite. Both these reaction fronts are sharp and are easily recognized in the field by their different colours. The mineralogical and chemical changes during the first stage are expressed by transformation of original oligoclase to albite, biotite (annite-rich) and hastingsite (amphibole) to hastingsite with low XFe values, dehydration, gain in Na, and losses in Fe and Rb. The second stage of metasomatism caused almost complete conversion of microcline to albite and complete or nearly complete disappearance of amphibole. Chemically, these changes are manifested by substantial gain in Na and extreme losses in K, Rb, Ba, Ca, Sr, Fe, and Mg. Depending on the modal abundances of amphibole, stage II albitized rocks are depleted in light rare earth elements or heavy rare earth elements or both, signifying that rare earth elements are principally hosted by mafic phases. The disparity in whole-rock δ18O values during both stages of albitization is related to the variations in modal amounts of Si-bearing phases. The replacement microstructures are in accord with the fluid-mediated phase transformations by a coupled dissolution–precipitation mechanism. The albitizing event took place at low temperatures of 350–400 °C and the fluid was metamorphic in nature.


Introduction
It has become increasingly clear that granitoid rocks can be hydrothermally altered beyond recognition on regional scales due to fluid-rock interactions (e.g. Plümper and Putnis 2009;Kaur et al. 2014 and references therein). The process of albitization is one of such common metasomatic phenomena. It is prevalent in varied tectonic settings, affecting a wide range of rock types (see Perez and Boles 2005 for a review), ranging in age from~3.0 to 0.3 Ga (e.g. Poujol et al. 2010;Jaguin et al. 2013). The process, in general, refers to the conversion of plagioclase and/or alkali feldspar into nearly pure albite by Na-rich fluid and has been recorded extensively in granitoids (e.g. Boulvais et al. 2007;Engvik et al. 2008;Kaur et al. 2012;among others). These studies also demonstrate that during such pseudomorphic replacement of feldspars by albite, the overall texture of the rock is retained in spite of the abrupt compositional changes in the rocks. The rate of phase transition by such a replacement mechanism is quite rapid, and it can transform the mineralogy of crustal rocks on a regional scale if sufficient fluid is available (Hövelmann et al. 2010;Norberg et al. 2011;Jamtveit and Hammer 2012). When such metasomatic replacements are expressed in the formation of sharp fronts and on a spatial scale exceeding metres, this indicates a significant role of infiltration mechanism for the chemical mass transport (Korzhinskii 1968). More important, at times, the reactive fluids can pervasively infiltrate a solid rock and transform the mineral assemblage to such an extent that it becomes a challenge to find parent-product pairs in the absence of a clear reaction front on the outcrop scale (Kaur et al. 2014). When the original protolith is no longer preserved, it is not possible to study the reaction mechanisms and metasomatic changes systematically. This is because it becomes impossible to have any control on the compositions and textures prior to metasomatism; thus eventually, the rocks may be assigned an incorrect origin (e.g. Elburg et al. 2001;Kaur et al. 2014).
Recognition of the hydrothermal alteration of granitoids on a regional scale can eliminate many misinterpretations regarding their characterization and origin as demonstrated for the granitoids of northern Aravalli orogen in Rajasthan, northwest India (Figure 1; Kaur et al. 2014). Recently, it has been shown that virtually all the 1.72-1.70 Ga ferroan granites of the Khetri complex of Rajasthan ( Figure 1) were albitized to varying extents (Kaur et al. 2011a(Kaur et al. , 2012(Kaur et al. , 2014 ca. 900 Ma after their emplacement, at around 850-830 Ma (Kaur et al. 2013a). A few ferroan intrusives of similar age also occur in the Alwar complex (e.g. Biju-Sekhar et al. 2003), located about 95 km SE of Khetri. Of these, only the Ajitgarh pluton has been subjected to detailed geochronological and geochemical studies (Pandit et al. 1996;Khatatneh 1998, 2003;Biju-Sekhar et al. 2002). These rocks are interpreted to represent magmatic mineralogy, consisting of a leucocratic trondhjemite/low-K granite in association with an alkali granite and both are ascribed to represent different pulses of anorogenic magmatism, derived by low-degree partial melting of an amphibolite source Khatatneh 1998, 2003). Favouring a magmatic origin for low-K granite, an alkali exchange mechanism across the interface between granitoid and basic magma has been suggested, whereby removal of K from the granite melt formed the low-K granite (Pandit and Khatatneh 2003). Nevertheless, a reassessment of the field and chemical relationship given in the present work shows that the Ajitgarh intrusive is also variably albitized. The different extents of albitization are not only easily recognized on the outcrop scale by the occurrence of two discrete metasomatic zones, but also are demarcated distinctly by two sharp reaction fronts. These reaction fronts mark the two stages of albitization in accord with the model outlined by Kaur et al. (2012Kaur et al. ( , 2014 for the northern Khetri ferroan granites. In the latter, however, a very small volume of unaltered original granite is preserved in merely two intrusives due to the pervasive nature of metasomatism in the region. This has resulted in the incomplete assessment of metasomatic chemical changes and also hampered the robust magmatic characterization of such granites. In contrast, a relatively large extent of original granite is exposed at Ajitgarh along with two distinct and sharp metasomatic fronts. This makes the intrusive an ideal subject for improving our understanding of the systematics of the two-stage metasomatism in granitoids. Specifically, this study focuses on the new field, mineralogical, geochemical and isotope data of the Ajitgarh intrusive of the Alwar complex, and these data are further compared with those of two other major granitoid plutons of the southern Khetri complex. The results are interpreted and discussed in the framework of previously published data for the northern Khetri ferroan granites.

Geological framework Regional geology
Two Proterozoic cover sequences, the Aravalli fold belt and the Delhi fold belt, which rest unconformably over an Archaean basement (Figure 1; e.g. Sinha- Roy et al. 1998;Roy and Jakhar 2002), sum up the Precambrian geology of the Aravalli orogen. The Archaean nucleus in the southern Aravalli region is known as the Banded Gneissic Complex (Heron 1953). It is dominated by a late Palaeoarchaean (~3.3 Ga) ensemble of tonalite-trondhjemite-granodiorite gneisses with minor enclaves of amphibolites and metasediments, and intruded by Neoarchaean (~2.8-2.5 Ga) granitoids (Gopalan et al. 1990; Wiedenbeck and Goswami 1994;Roy and Kröner 1996;Wiedenbeck et al. 1996). The Archaean basement containing numerous Palaeoproterozoic granulite bodies in the central part of the Aravalli orogen is referred to as the Sandmata complex (Roy et al. 2012). Although, these basement rocks are lithologically similar to the Palaeoarchaean component of the Banded Gneissic Complex, they span an age range of 2.9-2.5 Ga (Dharma Rao et al. 2011;Roy et al. 2012). The age of the tectonically emplaced granulites, within this Archaean gneiss-amphibolite-metasediment terrane, is in the range of~1.73-1.62 Ga (e.g. Roy et al. 2005;Buick et al. 2006;Bhowmik et al. 2010). The Sandmata complex also records an imprint of a collisional orogeny during Grenvillian times at around 1.0-0.9 Ga (Bhowmik et al. 2010).
The oldest cover sequence to the Archaean basement rocks is the Aravalli fold belt; it essentially comprises (meta)sedimentary rocks with a basal mafic volcanic sequence that is around 2.3-1.8 Ga in age (Ahmad et al. 2008). The rocks of the Aravalli fold belt around Udaipur show two metamorphosed, sedimentary facies associations with a carbonate-dominated near-shore shelf facies in the eastern part and a mudstone-dominated deepsea facies in the western part (e.g. Roy and Paliwal 1981). These sedimentary facies are divided into three lithostratigraphic units as the lower, middle and upper Aravalli Groups, each separated by an unconformity (Roy 2000). The depositional age of these units range from~2.5 (?) to 1.6 Ga (McKenzie et al. 2013). The Delhi fold belt is the younger cover sequence that is exposed in the northern and western parts of the Aravalli orogen ( Figure 1). It is dominantly a pelitic-psammitic (meta)sedimentary succession with minor carbonate and mafic components. The belt is further separated into two domains (Sinha-Roy 1984), mainly on the basis of divergent geochronological data obtained from granitoid rocks. The granitoids north of Ajmer are older (~1.85-1.70 Ga), whereas younger granitoids (~1.0-0.85 Ga) are exposed in the region south of Ajmer (Kaur et al. 2011b and references therein). The Delhi sediments in the northern domain were deposited at <1.7 Ga (Kaur et al. 2011b), whereas in the south, the same were deposited between 1.24 and 0.86 Ga (Singh et al. 2010;McKenzie et al. 2013).
The present work is focused on the northern domain of the Delhi fold belt (Figure 1), where 1.85-1.7 Ga granitoids form the basement for the rocks of Delhi Supergroup (Biju-Sekhar et al. 2003;Kaur et al. 2011b). The latter is divided into three lithological units: the lower Raialo Group (dominantly calcareous), followed by the Alwar Group (dominantly arenaceous) and the upper Ajabgarh Group (dominantly argillaceous). The northern Delhi fold belt is divisible into three igneous-metamorphic complexes, which from east to west are: Lalsot-Bayana, Alwar and Khetri. The present work is confined to the ferroan granites of the Alwar and Khetri complexes.

Local geology
The Ajitgarh intrusive is located in the Alwar complex and crops out amidst Quaternary sand and alluvium (Figures 1  and 2B). The ridges exposed to the east of Ajitgarh belong to the rocks of Alwar Group, which constitute sericite quartzite with amphibolite bands and lenses of metaconglomerate ( Figure 2A). This outcrop pattern with a 'hookshaped' geometry in the north represents a very large isoclinal anticline that is cut by a NW-SE-striking fault (Das 1988). It is an irregular-shaped pluton and constitutes three types of granites ( Figure 2B; nomenclature of granites after Kaur et al. 2014 and references therein). The northern part of the intrusive, is composed of unaltered original granite (Figure 3; inset A), which is grey and foliated in a NE-SW direction with westerly dips. The previous workers, however, described these granites as massive and non-foliated (Pandit and Khatatneh 1998). This is followed by a non-foliated pink microcline-albite granite facies towards the southwestern part of the pluton ( Figures 2B and 3; inset B), which represents the moderately albitized granite. The contact between both these granite types is sharp and at the contact, the pink granite  Heron (1923), Das (1988), and Singh (1988), whereas the portion south of Jaipur is from Roy and Jakhar (2002), Roy et al. (2012)). occurs as enclave-like bodies within the grey granite ( Figure 2C). At the southwestern extremity of the intrusive, a white to light grey, non-foliated albite granite is exposed that shows a sharp contact with the pink granite and also contains the enclaves of the latter (Figures 2B, 2D and 3; inset C); it represents the completely albitized granite. These granite facies are best exposed along the southwestern margin of the intrusive (Figures 2 and 3). Veins of pegmatite and quartz are observed in the microcline-albite and albite granites, mostly in the southwestern part of the intrusive (also see Pandit and Khatatneh 1998). Electron probe micro-analyser (EPMA) chemical dating of zircon grains from all the three granite facies shows an age range of 1.74-1.72 Ga (Biju-Sekhar et al. 2002). The Udaipurwati intrusive is located within rocks of the Alwar Group in the southern domain of the Khetri complex ( Figure 4). It is an elliptical pluton that occupies the axial zone of major antiform of quartzite and metapelitic schist (Das Gupta 1968;Gupta et al. 1998). The intrusive is largely composed of a well-foliated, grey to pinkish grey original granite ( Figure 5A). The microcline-albite occurs as enclave-like bodies within the original granite, in the northern and southern part of the intrusive, and is a foliated, pink to greyish pink rock. The grey to occasionally pinkish grey, foliated albite granite is of minor occurrence and is confined   to the northeastern and southwestern margins of the intrusive. These granites yielded EPMA chemical zircon ages of 1.7-1.68 Ga (Biju-Sekhar et al. 2003) and U-Pb zircon vapour transfer age of ca. 1.69 Ga (Kaur et al. 2011a).
The Mandaora pluton is also located in the southern Khetri complex, within the rocks of the Alwar Group, about 8 km NE of the Udaipurwati intrusive ( Figure 4). It is an oval-shaped intrusive that occurs in the core of an  antiform of quartzite (Gupta et al. 1998). A zone of metaconglomerate occurs along its northern and western margin ( Figure 5B), which largely shows deformed pebbles set in a matrix of quartzite ( Figures 5C and 5D). This provides evidence for the basement nature of the intrusive. Moreover, a thin mylonitic metapelite zone occurs between the conglomerate and the granite. The intrusive is largely made up of grey to white, foliated albite granite with only a small relic of megascopically similar microcline-albite granite in the central part of the granite body. These granites provided a thermal ionization mass spectrometry multigrain U-Pb zircon age of ca. 1.68 Ga (Sivaraman and Raval 1995), which is indistinguishable from the U-Pb zircon vapour transfer age of ca. 1.68 Ga (Kaur et al. 2011a).

Analytical techniques
Most of the analyses of major rock-forming minerals were carried out at the Wadia Institute of Himalayan Geology, Dehra Dun, and some at the Max-Planck-Institut für Chemie, Mainz (JEOL Superprobe JXA-8200). At Dehra Dun, a SX100 CAMECA EPMA was used under comparable operating conditions of 15 KeV accelerating voltage, 20 nA beam current, and beam size of 1 μm. The standards and the crystals used during the analysis were orthoclase (TAP), kyanite (TAP), TiO 2 (PET), Almandine (LIF), diopside (TAP), rhodonite (LIF), wollastonite (PET), jadeite (TAP), orthoclase (PET), and chromite (LIF). The PAP correction program after Pouchou and Pichoir (1985) has been applied to the mineral data. The analytical details for Mainz EPMA have been given in Kaur et al. (2006). The mineral symbols used are those of Whitney and Evans (2010).
Whole-rock major and trace element analyses were obtained from Activation Laboratories Ltd., Ontario. Major and trace elements were analysed by inductively coupled plasma optical emission spectrometry and inductively coupled plasma mass spectrometry, respectively, both utilizing the lithium tetraborate fusion technique (for detail see www.actlabs.com). Nd-Sr isotopic analyses were carried out at the Max-Planck-Institut für Chemie in Mainz as per the analytical details given in Kaur et al. (2011a). The Nd and Sr isotope ratios were determined on a TRITON Thermo, whereas Sm and Rb were measured on a Finnigan MAT 261 thermal ionization mass spectrometer equipped with a multi-collector. The oxygen isotope compositions for the whole-rock samples were measured using the BrF 5 method of Clayton and Mayeda (1963) on a duel inlet Finnigan MAT252 isotope ratio mass spectrometer at the Activation Laboratories Ltd., Ontario, as per the method detailed at www.actlabs. com. Isotopic compositions are reported relative to VSMOW, and the analyses were reproducible to ±0.2‰. The in-house standard African sand quartz (δ 18 O = 9.9‰) was used, and it has been calibrated against the NBS-19 standard.

Petrography Ajitgarh
The original granite is generally coarse grained (≥3 mm), with a subhedral granular microstructure ( Figure 6a). Quartz, microcline, plagioclase, biotite and amphibole are the major rock-forming phases, while titanite, apatite, zircon and epidote (rare) represent accessory minerals. Quartz is mostly present as subhedral, equigranular and elongated crystals, which at times show the formation of subgrains. Microcline (Or 96.2 ± 0.4 , n = 15; Table 1) generally occurs as subhedral tabular crystals and occasionally shows microperthite texture. Subhedral elongated plagioclase commonly shows albite twinning, and a few crystals display turbid/cloudy cores, sericitization and myrmekite intergrowths. The mineral is oligoclase in composition with an average An 15.5 ± 4.5 (n = 25; Table 2). Dark green, prismatic amphibole is hastingsite ( Figure 7A) with X Fe [= Fe t /(Fe t + Mg)] values between 0.87 and 0.91 and the Cl concentrations of around 2.47 wt% (Table 3). Biotite commonly shows a peritectic reaction relationship with amphibole ( Figure 6b). The mineral is rich in FeO t and poor in MgO with high and uniform X Fe values (0.75-0.77; Table 4), conforming to an annite-rich biotite ( Figure 7B).
The microcline-albite granite is relatively less coarse grained and locally shows a porphyritic texture. The essential rock-forming felsic phases in the rock are plagioclase, quartz and microcline, while amphibole is the main mafic phase and biotite is absent. Accessories are the same as in the original granite with the only difference being that epidote is more common in occurrence. The relative decrease in grain size is mainly related to the reduced size of microcline crystals due to its incipient alteration to albite ( Figure 6C), although a few megacrysts of the mineral are still preserved. The crystals of quartz are commonly equigranular, but the frequent presence of coarser recrystallized grains imparts a porphyritic-like texture to the rock ( Figure 6D). Compositionally, the microcline (Or 96.2 ± 0.6 , n = 14; Table 1) is identical to that of the original granite. Plagioclase in this rock is often cloudy, sericitized, and epidotized. Significantly, the mineral is almost pure albite with an average An 2.3 ± 1.2 (n = 21; Table 2). Amphibole varies in composition from hastingsite to ferro-edenite ( Figure 7A). It shows relatively low and more varied X Fe values (0.65-0.83) and also lower Cl contents (1.81 wt%) compared with the hastingsite of the original granite (Table 3).
The albite-granite shows similar grain size to that of microcline-albite granite. The rock is essentially composed of quartz and albite (An 0.7 ± 0.4 , n = 20) with minor to negligible amphibole. Pandit et al. (1996), however, reported oligoclase (An 14-24 ; Michel-Levy method) in this granite. Calcite appears as an additional accessory mineral, while epidote is not observed in the rock. Also the amounts of titanite and apatite are lower than that in the original and the microcline-albite granites. The albite is optically clear and sometimes also shows pericline twinning in addition to common albite twinning. Occasional chessboard-twinned tabular albite is also present, which is considered to have formed by complete replacement of K-feldspar by Na-feldspar (Kaur et al. 2014 and references therein). Even though the rock is non-foliated due to lack of biotite, the slight offsetting in twin lamellae of albite indicates that the original rock was deformed prior to albitization.

Udaipurwati
The grey original granite is medium grained and is mostly porphyritic due to the presence of white phenocrysts of feldspars, whereas the pinkish grey variety is largely nonporphyritic. The major mineral constituents include quartz, microcline, plagioclase, biotite, and minor amphibole,   while the accessories are mainly titanite, zircon, apatite, and epidote. Quartz frequently shows undulose extinction, and both feldspars are generally clouded. Microcline (Or 95.0 ± 1.2 , n = 33; Table 1) commonly occurs as tabular/equant megacrysts and occasionally shows patchy perthite texture. Plagioclase is generally medium grained, elongated, sericitized and epidotized. The mineral is mostly oligoclase with An 10.6-19.5 (An 14.4 ± 2.2 , n = 28; Table 2), but is also occasionally albite with An 4.5-9.7 . Amphibole is hastingsite in composition with average X Fe value (0.89) almost identical to the original granite of Ajitgarh ( Figure 7A), but with lower average Cl value (1.25 wt%). Brown biotite, relatively more abundant and common than amphibole, defines the main foliation. It is annite rich with higher X Fe values (0.79-0.81) than those of the Ajitgarh counterpart ( Figure 7B).
The microcline-albite granite is medium grained with a smaller amount of mafic minerals than in the original granite. Quartz, microcline, plagioclase and biotite constitute the major rock-forming phases accompanied by accessory zircon, apatite and titanite. Overall, the crystals are relatively more stretched and deformed than in the original granite. Although megacrysts of microcline (Or 96.1 ± 0.4 , n = 8) are present, the mineral also occurs as a relict phase within the megacrysts of plagioclase and at places is also completely altered to albite. Moreover, the amount of plagioclase in this rock is higher than in the original granite, but the mineral here is exceptionally pure albite with An 2.2-3.4 (An 2.8 ± 0.5 , n = 5). At a few places, however, the magmatic composition of plagioclase is still preserved  ), similar to that in the original granite. In addition to medium-sized quartz, as observed in the  (Table 4 and Figure 7B).
The albite granite is medium grained and is characterized by almost complete absence of microcline, although relics of microcline occur sporadically. Hence, it is essentially a biotite and quartz-bearing albite granite. Muscovite is present as an additional accessory phase, while the other accessory phases, such as titanite and apatite, are almost negligible in some samples. Chessboard albite is prominently developed in the rock ( Figure 6E), and bending of twin lamellae is common along with occasional sericitization in the albite (An 3.7 ± 0.7 , n = 11; Table 2). Biotite is annite-rich, but with much lower X Fe values (avg. 0.60) compared with the original and microcline-albite granites ( Figure 7B).

Mandaora
The major mineral constituents of microcline-albite granite are represented by quartz, microcline, plagioclase, and biotite. Microcline (Or 94.2 ± 0.7 , n = 7) mostly occurs as tabular megacrysts and often exhibits irregular and diffuse boundaries. Plagioclase is generally tabular in shape and is mostly sericitized. The mineral is severely albitized resulting in composition An 2.1-3.6 (An 2.9 ± 0.6 , n = 10), nearly identical to the Udaipurwati counterpart (Table 2). Biotite is annite (avg. X Fe = 0.74) but is relatively Al-rich compared with the biotites of Ajitgarh and Udaipurwati (Table 4 and Figure 7B). The albite granite, like in other intrusives, is devoid of microcline and is essentially a albite-quartz rock with biotite and/or amphibole as major mafic phases. In some samples (e.g. MD-8), secondary muscovite is abundant along with accessory minerals titanite, apatite, and zircon. Exceptionally pure megacrysts of albite (An 1.8 ± 0.2 , n = 19), sometimes with deformed twin lamellae, commonly display combined chessboard and Carlsbad twinning ( Figure 6F). Biotite, as in Udaipurwati albite granite, is relatively poor in Fe and rich in Mg, resulting in low X Fe values (avg. 0.58; Figure 7B).

Elemental geochemistry
The original granite at Ajitgarh shows an evolved (SiO 2 = 70.5-72.7 wt%), but a relatively uniform chemical composition for most of the major and trace elements ( Table 5), indicating that the granites were not variably fractionated. The same is also true for the original granite of Udaipurwati (Table 6) although it seems to be somewhat more evolved (SiO 2 = 70.8-73.7 wt%; excluding UW-21) than that of Ajitgarh. The Si contents in the microcline-albite granite (SiO 2 = 71.7-72.5 wt%) of Ajitgarh are similar to that of the original granite, but these are significantly higher in the albite granites (SiO 2 = 74-76 wt%; Table 5). On the other hand, such a systematic increase in Si with increasing extent of albitization is not observed in the Udaipurwati intrusive, but at Mandaora, some of the albite granites are relatively enriched in Si compared with the microcline-albite granite ( Table 7). The albite granites can be readily distinguished from the original and the microcline-albite granites by a remarkable increase in Na concentrations (as high as 7 wt%) and extremely low K (as low as 0.10 wt%), Rb, Ba, and Fe (sometimes) abundances (Tables 5-7). An extreme drop in Rb from 285 ppm (in the original granite) to merely 2 ppm (in the albite granite) with increasing Na/K, taken as an index of albitization, is discernible in the albitized granites studied (Figure 8). Some of the albite granites of the northern Khetri complex even show Rb <1 ppm. Also, the nature of mafic phases controls the Rb values the original granites, for example, the biotite-rich    Udaipurwati granites show relatively high Rb values compared with the amphibole-rich Ajitgarh granites. In comparison to the previously published data on the Ajitgarh granites (Pandit et al. 1996;Pandit and Khatatneh 1998;Figure 8), the 'alkali granites' described by these authors show similar Na/K ratios and corresponding Rb values to the original and microcline-albite granites (this study). In contrast, the 'low-K granite/trondhjemite' reported by Pandit and co-workers (Pandit et al. 1996;Pandit and Khatatneh 1998), compared with mineralogically equivalent albite granite of the present study, shows significantly lower Na/K ratios (Na/K =~4; Na 2 O =~4 wt% and K 2 O =~1.5 wt%;  Table 5) as well as those of the northern Khetri complex (Figure 8). This is further corroborated by the fact that the only K-bearing phase in the albite granite is amphibole, which occurs in minor or negligible amounts. Relatively high K 2 O and low Na 2 O values of these samples drive them towards the lower Na/K ratios, thereby developing a pseudo-field for low-K granite/trondhjemite (Figure 8). We do not know the reasons for these discrepancies. Conceivably, they may reflect analytical error in the major element data (alkali elements, in particular) published in Pandit et al. (1996) and Pandit and Khatatneh (1998).
In general, the U and Th values are lower in the Ajitgarh intrusive compared with Udaipurwati and Mandaora. It should be noted that the measured U values in most of the granites are variable and low, leading to inconsistent and high Th/U ratios relative to the normal crustal Th/U ratio (~4; Rudnick and Gao 2003), and they are unlikely to be primary. This is most likely due to the removal of U by oxidation and dissolution by ground water and weathering in recent times. Therefore, the U content can be assumed here to be higher than the measured U values, and in order to obtain more realistic present-day internal heat production values for the granites, the U concentrations must be increased by a factor (1.5-1.7) to get the normal crustal Th/U, thereby compensating for U removal in recent times. Excluding albite granites (because of loss of K), the internal heat production values for Ajitgarh granites (avg. 3.5 μWm −3 ), are lower than Udaipurwati (avg. 6.3 μWm −3 ) and Mandaora (5.5 μW m −3 ; Tables  5-7). On average, granite shows the median heat production value of 2.43 ± (1.74 − 3.23) μW m −3 (Vilà et al. 2010). The studied granites are thus considered as high in heat production.
The total rare earth element (REE) contents of the albite granites in the Ajitgarh intrusive are drastically lower than the original and microcline-albite granites; however, such a depletion in REE is not observed in the case of the Udaipurwati and Mandaora (except for one sample, MD-8) intrusives (Tables 5-7). The original and microcline-albite granites in Ajitgarh show similar and fractionated [(La/Yb) N = 5.8-7.0] REE patterns with nearly flat heavy REE profiles and prominent negative Eu anomalies ( Figure 9A and Table 5). The REE patterns of the albite granites are, however, relatively less fractionated because of loss in light REEs. Moreover, the samples that are devoid of any mafic phases show the lowest REE abundances, for example, AJ-20 ( Figure 9A). The negative Pr and Er anomalies reported by Pandit and Khatatneh (2003) in the Ajitgarh granites are not observed in our samples. In particular, the extremely low Pr values are highly unusual, and we are not aware of other granites displaying such a REE pattern. The REE patterns for all the granite types of the Udaipurwati intrusive are rather similar ( Figure 9B), but are more and variably fractionated [(La/Yb) N = 8.8-28.3] than that of Ajitgarh. The Mandaora albitized granites also show similar REE profiles ( Figure 9C), but relatively less fractionated patterns [(La/Yb) N = 6.5-13.1; excluding MD-8] than that of Udaipurwati. The primitive mantle-normalized multi-element patterns of all the intrusives studied are characterized  Chaudhri et al. (2003) and Kaur et al. (2006Kaur et al. ( , 2012Kaur et al. ( , 2014, and those of Ajitgarh alkali granite and low-K granite are from Pandit and Khatatneh (1998).
by prominent negative anomalies in K, Ba, Sr, P, Eu, and Ti, which are relatively more pronounced in the albite granites, especially those of Ajitgarh ( Figures 9D-F). Even though metasomatized rocks are not suitable for magmatic classification and discrimination diagrams, Kaur et al. (2014) demonstrated that the moderately albitized microcline-albite granite can be correctly characterized, whereas this is not true for the completely albitized albite granites. Therefore, the albite granite samples are excluded for magmatic characterization, and only those classification diagrams are used which are based on relatively immobile major and trace elements. The ASI [molar     Figure 10A). In the Fe number (Fe*) classification diagram of Frost et al. (2001), the majority of the samples are ferroan (high-Fe*) in character and occupy the field of A-type ferroan granites ( Figure 10B). The Zr-Ga/Al classification diagram of Whalen et al. (1987) also confirms the A-type affinity of  et al. (1984), (E) and (F) representative discriminant function multi-dimensional discrimination diagrams based on ln-transformed ratios of immobile major and trace elements for island arc (IA), continental arc (CA), continental rift (CR), ocean-island (OI), and collision (Col) tectonic settings; the subscript 'mtacid' indicates major and trace element ratios (after . Plot symbols and data sources are the same as in Figure 8. the studied granites along with those of the northern Khetri complex granites ( Figure 10C). In the Nb-Y tectonic discrimination diagram of Pearce et al. (1984), the samples plot in the overlapping region of within-plate and ocean ridge granites (dashed line in Figure 10D). In order to confirm the tectonic setting of these granites, we have used new multi-dimensional diagrams based on log ratio transformations proposed by Verma et al. (2012. The major element data were adjusted after Fe-oxidation adjustment on an anhydrous basis to 100%, using software IgRoCS (Verma and Rivera-Gómez 2013). In the first set of diagrams, based on log-ratios of major elements ( Figure S1; Verma et al. 2012; for online supplemental figures, please see http://dx.doi.org/10.1080/ 00206814.2014.1000394), no firm interpretation of the tectonic environment can be made as the samples straddle the boundary between collision and continental rift (Figures S1A, S1D and S1E) or continental arc and continental rift ( Figure S1B) or they even plot in the collision field ( Figure S1C). The same results are also obtained in the next set of five diagrams based on log ratios of immobile trace elements ( Figure S2; ). However, consistent results are obtained in another set of five diagrams, which are based on log-ratios of immobile major and trace elements ( Figure 10E and F; . These diagrams support the within-plate tectonic setting as all samples (100%) plot in the combined field of continental rift and ocean island (for full set of diagrams, see Figure S3).

Isotope geochemistry
The Sm-Nd isotope system (Table 8) bears no geochronological information as the granites under consideration when plotted as 'isochron' (ISOPLOT, Ludwig 2012) show a similar steep slope (~2.6 Ga; Figure 11), as defined by the northern Khetri granites with 147 Sm/ 144 Nd < 0.15 (Kaur et al. 2014). A trend corresponding to the slope of 1.4 Ga was obtained for the granites having higher 147 Sm/ 144 Nd (>0.15) but too low 143 Nd/ 144 Nd for 2.6 Ga. It is noteworthy that such granites have lost nearly all their mafic phases (amphibole) during the advanced stage of albitization accompanied by depletion in LREE.
In view of high mobility of both Rb and Sr during metasomatism, the Rb-Sr system in the present set of samples is highly perturbed as reflected by unrealistically low initial Sr values, when calculated for an emplacement and closure age of 1.7 Ga (Table 8). The Rb-Sr system was severely disturbed because the minerals which control the abundances of Rb (K-feldspar and biotite) and Sr (plagioclase and amphibole) were subjected to varied extents of metasomatism, thereby yielding meaningless information due to differential loss of Rb and Sr. This disturbance must have taken place at a time much later than 1.7 Ga. The whole-rock δ 18 O values for the microcline-albite granites in Ajitgarh and Udaipurwati are relatively low compared with their original granites (Table 8), but these are still within the normal range found in granites (+6 to +10‰; Taylor 1978). In the case of the northern Khetri albitized granites, these values show either an increase or no change in whole-rock δ 18 O values during the first stage of albitization, that is, the formation of microcline-albite granite from the original granite ( Figure 12; see also Kaur et al. 2012Kaur et al. , 2014. Nearly, all the albite granites have higher δ 18 O values than the microcline-albite granites, except for the Mandaora albite granites, which have somewhat lower δ 18 O values. In general, there is enrichment in whole-rock δ 18 O values during the extreme stage of albitization, i.e. the formation of albite granite from microcline-albite granite. It should be noted that no generalized trend is discernible between the whole-rock δ 18 O values and the extent of albitization/silica enrichment for the granites under consideration, unlike the increasing trend observed for the albitized granites previously studied ( Figure 12).

Discussion
Two distinct sharp reaction fronts can be readily recognized on the outcrop scale in the Ajitgarh pluton. The first reaction interface is noticed between the protolith (original granite) and stage I metasomatized rock (microcline-albite granite). The colour transition from grey for original granite to pink for the microcline-albite granite marks this reaction front ( Figure 2C). The second reaction interface is distinctly expressed megascopically by the change in colour from pink microcline-albite granite (stage I) to stage II white albite granite ( Figure 2D). Furthermore, these reaction fronts are also marked by the abrupt change in mineralogy accompanied by abrupt compositional and isotopic changes. By contrast, none of the reaction fronts are noticed in the field at the Udaipurwati and Mandaora intrusives, although the same can be observed for some of the northern Khetri granites (see Kaur et al. 2012).
The major-and trace-element changes during the two stages of a metasomatic event have been assessed in isocon diagrams (Figures 13, S4 and S5), following the approach of Grant (2005) as detailed in Kaur et al. (2012). Here, a reference line corresponding to a zero concentration change (an isocon) intersects the axes at the origin of the plot (Grant 1986). The slope of the isocon may be based on constant concentration of some component (e.g. Al), or of mass, or of volume. Assuming that there was no change in volume during albitization, isocons with a slope of 1 are drawn (e.g. Boulvais et al. 2007).

Mineral transformations and chemical changes across reaction front-I
At Ajitgarh, the transition from grey to pink colour across reaction front-I is mainly due to reddening of K-feldspar. The latter appears red because of the formation of haematite dust during hydrothermal alteration under oxidizing conditions because Fe 3+ , which is tetrahedrally incorporated in the structure of K-feldspar, forms haematite dust on oxidation (e.g. Plümper and Putnis 2009). This front is characterized by nearly complete transformation of oligoclase to albite, incipient alteration of microcline to albite, hastingsite amphibole (X Fe = 0.88-0.89), and annite-rich biotite (X Fe = 0.75-0.77) to relatively Fe-poor, Cl-poor, Mg-rich, and Na-rich hastingsite amphibole (X Fe = 0.70-0.82). The compositional variation in the secondary amphibole is principally controlled by the fluid activity ratio aCl − /aOH - (Kullerud and Erambert 1999). This is because the cationic composition of amphibole is dictated by crystal chemical constraints, induced on the crystal structure by Cl incorporated on the anion side (Kullerud 1996). Therefore, the wide range in X Fe of newly formed hastingsite is merely a reflection of local compositional gradients in the evolving fluid phase. Also the decrease in its Cl content during stage I is possibly due to the pronounced preference for strong partitioning of Cl into the fluid phase relative to the mineral during Clbearing fluid-mineral interaction (Kullerud et al. 2001). A detailed and systematic study of relative losses, gains, and immobility of elements can be made during the stage I metasomatic alteration at Ajitgarh, which was previously not possible for other albitized granites of the northern Khetri complex because of limited and very small outcrops of unaltered original protolith. The chemical changes during metasomatism of the original granite (average of samples AJ-5, AJ-6, AJ-14, AJ-17, AJ-18) to the moderately albitized microcline-albite (average of samples AJ-3,  are shown in Figure 13A. This stage of metasomatism displays gain in Na, which is the outcome of transformation of oligoclase to albite. The Ca and Sr released during this transformation might have been partly fixed into hastingsite, and the accessory phases apatite and epidote, and thus these elements do not show any gain or loss. The almost complete transformation of original annite to hastingsite leads to loss in Fe and water [H for loss on ignition (LOI)]. The loss in Rb, but no such corresponding loss in K, is likely related to dissolution of annite because this mineral has higher Kd values for Rb than K-feldspar (Rollinson 1993).
The alteration trend of feldspars at Udaipurwati is similar to that at Ajitgarh, but in the former, there is no survival of hastingsite during the transformation of the original granite to the microcline-albite granite. The main mafic phase that continues to exist as well as additionally formed (from hastingsite) at this stage of  Figure 13. Isocon diagrams (after Grant 2005) showing the whole-rock chemical changes during the transformation of (A) the original granite to the microcline-albite granite (stage I metasomatism), (B) the microcline-albite granite to the albite granite with mafic minerals (stage II metasomatism), and (C) albite granite without mafic minerals (sample AJ-21) at the Ajitgarh intrusive. Major elements are given in wt% and trace elements in ppm, and labels for major elements are abbreviated to the elements. All elements are appropriately scaled to avoid crowding of data points (for detail, see Grant 2005;Kaur et al. 2012). metasomatism is annite-rich biotite. The isocon analysis for stage I between the original granite (average of samples UW-1, UW-2,  and the microcline-albite granite (average of samples  is shown in Figure S4A. In addition to gain in Na, as also observed at Ajitgarh, the losses in K, Ba, and Rb are related to the partial alteration of K-feldspar to albite. The Mg gain along with slight hydration and losses in Ca and Sr can be attributed to the breakdown of hastingsite to annite.
Overall, the changes during metasomatic reaction front-I seen in the Ajitgarh and Udaipurwati plutons are rather minor. In contrast with our earlier work on the northern Khetri granites (Kaur et al. 2012(Kaur et al. , 2014, in which the original granite is poorly represented, these new results are based on a solid data base for two bodies of original granites. Here, only two elements show consistent changes, namely gain of Na and loss of Rb. All other elements show either no significant change or apparently more random variations. As shown above, the whole-rock δ 18 O values do not show any direct correlation with either silica enrichment or with progressive albitization (Figure 12). In fact, δ 18 O values seem to be mainly related to the changes in the nature and the modal abundances of the additional Sibearing (re-)crystallized minerals and those being replaced or lost. If the silica enrichment is due to the (re-)crystallization of quartz accompanied by the dissolution of mafic phases (amphibole and/or biotite), then there is an increase in whole-rock δ 18 O values during stage I as observed in the Dosi intrusive of the northern Khetri complex (Kaur et al. 2012). This is due to the fact that, at any given temperature, quartz has the strongest tendency to concentrate 18 O among all the rock-forming minerals (Zheng 1993). Even though at Ajitgarh, there is evidence of recrystallization of strain-free quartz ( Figure 6D) during stage I, the whole-rock δ 18 O values decrease from the original to the microcline-albite granites ( Figure 12 and Table 8). This may be related to the reduced modal abundances of quartz in the microcline-albite granite. The replacement of oligoclase by isotopically heavy albite and biotite by heavy amphibole do not seem to have contributed much to the δ 18 O enrichment. In contrast, during stage I at Udaipurwati, there is a minor decrease in whole-rock δ 18 O values, which may be caused by the transformation of amphibole to biotite as the latter tends to be isotopically lighter than amphibole (Zheng 1993). This reflects that the whole-rock oxygen isotope compositions of the rocks during this stage are largely governed by modal abundances of those minerals that have been significantly influenced by the albitization.
The two-feldspar thermometer of Fuhrman and Lindsley (1988), when applied to the feldspars of microcline-albite granite formed at this stage (for detail on the aspect of equilibrium see Kaur et al. 2012 and references therein), yielded equilibrium temperatures (T ord ) of 340-350°C for the Ajitgarh and Udaipurwati intrusives, whereas the Mandaora microcline-albite granite provided a higher temperature of about 390°C, well within the range of albitization temperatures obtained for the granites of the northern Khetri complex (Kaur et al. 2012(Kaur et al. , 2014. On the other hand, it seems that the Ajitgarh and Udaipurwati intrusives were albitized at somewhat lower temperatures.

Mineral transformations and chemical changes across reaction front-II
The abrupt colour change from pink to white across the reaction front-II at Ajitgarh is mainly due to the whitening of feldspars and near-absence of mafic phases in the albite granites; this also causes destruction of foliation in the granites. The whitening of albite is related to the liberation of tetrahedrally incorporated Fe 3+ and Ti 4+ from the structure of K-feldspar when it is being replaced by albite (Norberg et al. 2011). The feldspar present here is essentially a virtually pure albite, and the rock is thus reduced to nearly a bimineralic (quartz and albite) granite. This is in accord with the Korzhinskii theory of infiltration mechanism, which tends to diminish the number of coexisting mineral phases (Korzhinskii 1968).
Two isocons are presented for the stage II metasomatism at Ajitgarh. The first isocon ( Figure 13B) considers those albite granite samples (average of AJ-9, AJ-11, and AJ-12) in which amphibole continues to exist, although in minor amounts. The isocon reveals a significant gain in Na and substantial losses in K, Rb, and Ba, which can be accounted for by the transformation of K-feldspar to albite. The losses in Ca, Sr, Fe, Mg, and water indicate their removal in solution, which is mainly related to reequilibration of amphibole during its interaction with fluid. This led to dissolution of amphibole as reflected by the reduction in its modal abundances. The losses in some of these elements are also reflected by relatively more pronounced negative anomalies in the multi-element patterns of albite granites ( Figure 9D). The complete to partial dissolution of accessory phases at this stage might also be partly responsible for certain losses; for example, epidote for Ca and H 2 O, apatite for Ca and P, and titanite for Ca and Ti. The dehydration may also be partly linked to the loss in sericite due to the formation of optically clear albite at this stage. The loss of La is noteworthy, and is mainly caused by the substantial loss of amphibole at this stage. In the second isocon ( Figure 13C), the albite granite sample (AJ-21) that is devoid of amphibole and nearly devoid of accessory phases is plotted against the same set of microcline-albite granite samples. It depicts more substantial losses in La, P, Fe, Mg, and Ca along with additional losses in Lu, Y, and Th. This signifies that the REEs are hosted almost completely by mafic phases, so that sample AJ-21, which has lost almost all of its Fe and Mg, has also lost most of the REEs. The dissolution of accessory phases might have also contributed to REE depletion, but not substantially. This is also reflected by the Sm-Nd isotope data (Figure 11), where high 147 Sm/ 144 Nd (>0.15) is observed only for those albite granites, which show relatively very low Nd concentrations, extreme REE depletion and almost complete loss of mafic minerals.
Besides gain in Na and losses in K, Rb, and Ba, related to the transformation of K-feldspar to albite, the isocon ( Figure S4B) shows that the rocks at Udaipurwati during stage II experienced gain in Mg and loss in Fe. This is primarily related to the difference in composition of biotite, which is relatively Fe poor and Mg rich in the Udaipurwati albite granite ( Table 4). The variation in the biotite composition is most probably the manifestation of dependency of Fe/Mg ratios of biotite on the Cl content of the externally derived fluid in equilibrium with biotite and the distribution of coexisting Ti-bearing minerals (Kullerud 1995). The dissolution of apatite seems to be responsible for losses in P, Ca, and Sr. The crystallization of non-sericitized new albite instead of sericitized feldspars in the microcline-albite granite seems to be accountable for the observed dehydration. The minor gain in La is most probably an artefact of magmatic fractionation (for detail, see Kaur et al. 2012). Similarly, at Mandoara, the gains in Mg and Na, and losses in K, Rb, Ba, Fe, Ca, P, and Sr ( Figure S5A) during transformation of the microcline-albite granite (sample MD-1) to the albite granite (average of samples MD-2, MD-3, MD-5, MD-6, MD-9 and MD-10) are related to similar mineralogical changes as in the Udaipurwati stage II metasomatism. Another sample of Mandaora albite granite (MD-8), in which most of the biotite is altered to muscovite, shows some additional gains and losses of elements ( Figure S5B). The most prominent gains are that of Al and Ca, and the losses are that of La, Lu, and Y. The Al and Mg gains and Fe loss are related to the formation of muscovite from biotite, whereas the prominent loss of La may also be related to the same mineralogical change, as muscovite is generally a poor carrier of LREE compared with biotite (Bea 1996). The Ca gain is related to the abundance of titanite and apatite in this sample.
There is a substantial increase in whole-rock δ 18 O values during the stage II relative to stage I, except for Mandaora (Table 8). Also at this stage, all the albite granites show higher modal abundances of albite formed by the replacement of K-feldspar and the breakdown of mafic phases (amphibole in Ajitgarh and biotite in Udaipurwati). The tendency of albite to concentrate 18 O is higher than K-feldspar and also much higher than the amphibole and biotite (Zheng 1993). Therefore, in all albite granites, there is an enrichment of whole-rock δ 18 O relative to microcline-albite granites. Although it is difficult to provide a sound explanation for the decrease of whole-rock δ 18 O values at Mandaora, the relative low modal abundances of quartz in the Mandaora albite granites seem to be the plausible reason.
A two-stage infiltration model and possible mechanism of metasomatism As the albitizing fluid metasomatized the granites at low temperatures well below any magmatic stage, then in principle, it may be either meteoric (δ 18 O = 0 to −10‰; Sheppard 1986) or metamorphic (δ 18 O = + 3 to +20‰; Sheppard 1986) in nature. The low temperature of albitization and the combined oxygen isotope data for the Alwar and Khetri albitized granites showing comparable (Ajitgarh, Udaipurwati and Tehara) or relatively enriched whole-rock δ 18 O values relative to the original granites confirm our previous contention for the metamorphic nature of the albitizing fluid (Kaur et al. 2012). This is because the meteoric water being very low in δ 18 O would rather cause depletion in whole-rock δ 18 O values by at least 6-12‰ (e.g. Taylor and Forester 1971).
The abrupt changes in the mineral assemblage across the two reaction interfaces (see above) demonstrate the formation of two sharp metasomatic replacement fronts, which are well displayed on the outcrop scale in the Ajitgarh intrusive due to the prominent colour differences. The presence of two sharp reaction fronts, produced by (relatively low-temperature) fluid infiltration, confirms the two-stage metasomatic model postulated for the albitization of the northern Khetri granites (for details, see Kaur et al. 2012). Briefly, the infiltration of low-temperature Na-rich fluid albitized the original granite in two discrete steps due to the difference in the rates of advancement of two reaction fronts. A leading metasomatic front transformed the original oligoclase into albite to form pink microcline-albite granite, followed by a more slowly moving second replacement front, which eliminated the K-feldspar and converted the microcline-albite granite into white albite granite. A schematic summary of the metasomatic process and the chemical changes induced by it is shown in Figure 14.
Recently, several experimental studies confirmed that the metasomatic replacement of feldspars by albite is a pseudomorphic phenomenon via an interface-coupled dissolution-precipitation mechanism (e.g. Putnis and Putnis 2007;Engvik et al. 2008;Hövelmann et al. 2010;Putnis and Austrheim 2010). Such a replacement is characterized by a sharp chemical interface between the parent and the product phase, preservation of crystallographic orientation, and external dimension of the parent phase. The continuation of twin lamellae of relict microcline within the replacing albite during stage I of the studied intrusives reflects the preservation of crystallographic orientation of the feldspars. The overall preservation of subhedral granular and graphic microtextures in the microcline-albite granites during stage I further points to isovolumetric replacement of feldspars. This also suggests the spatial coupling between the dissolution of initial parent phases and the precipitation of new product phases during this stage.
Nevertheless, the degree of spatial coupling seems to be reduced during stage II, specifically when there is formation of new albite at the expense of amphibole and/or biotite. The ultimate disappearance of ferromagnesian phases resulted in complete obliteration of (macroscopically visible) foliation from these rocks, which converted the initially foliated granite to massive albite granite. This may be due to the fact that the spatial coupling or separation between phases of dissolution and precipitation is dependent on the relative rates of dissolution, diffusive transport through the fluid and precipitation (Xia et al. 2009).
Based on the mineralogy and microstructures in conjunction with whole-rock geochemical and isotopic data, we infer that the formation of two distinct metasomatic zones (microcline bearing and microcline free) in the intrusives studied can be modelled as a result of only one event of influx of the externally derived Na-rich metamorphic fluid at temperatures of about 350-400°C.
The subsequent local compositional gradients in the fluid phase resulted in varied compositions of ferromagnesian phases. The high heat production of these ferroan granites might have caused long-term horizontal temperature gradients in the upper crust that served as heat pumps for driving circulation of hydrothermal fluids on a large scale.

Concluding remarks
(1) The two sharp replacement fronts in the ferroan granites of the intrusives studied are marked by prominent colour differences due to low-temperature (350-400°C) infiltration metasomatism in two simultaneous but spatially discrete steps, caused by a Na-rich metamorphic fluid.
(2) The reaction front I at Ajitgarh is characterized by nearly complete transformation of oligoclase to albite and incipient alteration of microcline to albite. Hastingsite amphibole and annite-rich biotite were replaced by relatively Fe-poor, Cl-poor, Mg-rich and Na-rich hastingsite amphibole. The decisive chemical changes during this stage of metasomatism include dehydration along with gain in Na and losses in Rb and Fe. In contrast, at Udaipurwati, both annite-rich biotite and hastingsite were transformed to biotite with lower X Fe . The relevant chemical changes include gain in Mg along with slight hydration and losses in Ca and Sr. (3) The reaction front-II at Ajitgarh is marked by nearly complete disappearance of microcline and amphibole. During stage II metasomatism, there is a significant gain in Na and substantial losses in K, Rb, Ba, and sometimes REEs accompanied by less extreme losses in Ca, Sr, Fe, Mg, and water. (4) The decreasing REE abundances and high 147 Sm/ 144 Nd (>0.15) in the albite granite mainly correlate with the reduced modal amounts of mafic minerals (largely amphibole) rather than the accessory phases. (5) The decrease or increase in whole-rock δ 18 O values is not directly related to Si enrichment but is mainly governed by the relative difference in modal abundances of Si-bearing phases. (6) The observed microstructures are in consonance with the features of interface-coupled dissolution-precipitation mechanism of metasomatic replacement.
IgRoCS and an unpublished version of the TecDIA software. His help in using this software is gratefully acknowledged. We express our gratitude to D. Rameshwar Rao for carefully performing EMP analyses and to Nusrat Eliyas for help rendered in the field work and sample preparation.