BADRINATH-GANGOTRI PLUTONS (GARHWAL, INDIA) - PETROLOGICAL AND GEOCHEMICAL EVIDENCE FOR FRACTIONATION PROCESSES IN A HIGH HIMALAYAN LEUCOGRANITE

The Gangotri leucogranite is the western end of the Badrinath granite, one of the largest bodies of the High Himalayan Leucogranite belt (HHL). It is a typical fine grained tourmaline + muscovite +/- biotite leucogranite. The petrography shows a lack of restitic phases. The inferred crystallization sequence is characterized by the early appearance of plagioclase, quartz and biotite and by the late crystallization of the K-feldspar. This suggests that, in spite of being of near minimum melt composition, the granite probably had long crystallization or melting interval, in agreement with previous experimental studies. Tourmaline and muscovite have a mainly magmatic origin. Even though the major element composition is homogeneous, there are several geochemical trends (when CaO decreases there is an increase in Na2O, Rb, Sn, U, B, F and a decrease in K2O, Fe2O3, TiO2, Sr, Ba, Zr, REE, Th) which are best explained by a fractionation process with early crystallizing phases. Experimental solubility models for zircon and monazite in felsic melt support a magmatic origin for these two accessory phases as well. Rb/Sr isotope data show this granite to have, like other HHL, heterogeneous isotopic values for Sr (initial Sr-87/Sr-86 ratios, calculated at 20 Ma, range between 0.765 and 0.785). Therefore no mixing (i.e. no convection) occurred between the different batches of magma. In contrast O-18 data show little variation (13.04 part per thousand +/- 0.25), implying a source with homogeneous O-18 values. Differences in timing between fluid infiltration and the onset of melting, related to differences in temperature of the source, could explain why source homogenization occurred for the Gangotri and not for the Manaslu granite. The use of experimental results for solubility and the position of the accessory minerals during melting, predict a low viscosity for the melt during its extraction. This in turn explains the lack or restitic phases (major and accessory) in the granite as well as some field features (lensoid shape, pronounced magmatic layering). Based on the petrographic and isotopic studies, it is suggested that the mechanism of ascent was not diapiric but rather that the melt ascended along several fractures and the level of emplacement was partialy controlled by the density contrast between the melt and host rocks.