The petrology and timing of crustal melting has been investigated in the migmatites of the Higher Himalayan Crystalline (HHC) exposed in Sikkim, India. The metapelites underwent pervasive partial melting through hydrous as well as dehydration melting reactions involving muscovite and biotite to produce a main assemblage of quartz, K-feldspar, plagioclase, biotite, garnet ± sillimanite. Peak metamorphic conditions were 8-9 kbar and *800 °C. Monazite and zircon crystals in several migmatites collected along a N-S transect show multiple growth domains. The domains were analyzed by microbeam techniques for age (SHRIMP) and trace element composition (LA-ICP-MS) to relate ages to conditions of formation. Monazite preserves the best record of metamorphism with domains that have different zoning pattern, composition and age. Zircon was generally less reactive than monazite, with metamorphic growth zones preserved in only a few samples. The growth of accessory minerals in the presence of melt was episodic in the interval between 31 and 17 Ma, but widespread and diachronous across samples. Systematic variations in the chemical composition of the dated mineral zones (HREE content and negative Eu anomaly) are related to the variation in garnet and K-feldspar abundances, respectively, and thus to metamorphic reactions and P-T stages. In turn, this allows prograde versus decompressional and retrograde melt production to be timed. A hierarchy of timescales characterizes melting which occurred over a period of *15 Ma (31-17 Ma): a given block within this region traversed the field of melting in 5-7 Ma, whereas individual melting reactions lasted for time durations below, or approaching, the resolution of microbeam dating techniques (*0.6 Ma). An older *36 Ma high-grade event is recorded in an allocthonous relict related to mafic lenses. We identify two sections of the HHC in Sikkim that traversed similar P-T conditions at different times, separated by a tectonic discontinuity. The higher structural levels reached melting and peak conditions later (*26-23 Ma) than the lower structural levels (*31-27 Ma). Diachronicity across the HHC cannot be reconciled with channel flow models in their simplest form, as it requires two similar high-grade sections to move independently during collision.

No abstract

Abstract. Fe-Mg interdiffusion rates have been measured in olivine solid solutions using the diffusion couple technique. Measured diffusion rates along [001] at fo2 = 10 -•2 bars and between 980øC and 1300øC are found to be slower by about 2 orders of magnitude, compared to previous studies with the exception of some isolated data points of Misener [1974]. There is no change in temperature dependence within this temperature interval for any composition studied. The temperature dependence at a composition of Fo86 is described by an activation energy of 226+18 kJ/mol and a preexponential factor of (5.38+0.89)x10 -9 m2/s. The data are consistent internally as well as with well-established theoretical models relating tracer and chemical diffusivities. It is shown that diffusion coefficients are a unique function of pressure, temperature, major element composition and oxygen fugacity for olivines and are not dependent on trace element contents of starting crystals. It is argued that a twofold classification of diffusion mechanisms into intrinsic and extrinsic is inadequate for Fe-bearing silicates, and at least three categories need to be defined. Consideration of the data in combination with the point defect chemistry indicates that the observed diffusion in olivines occurs within a "transition-metal extrinsic" regime. For low temperature extrapolations, significant changes in temperature dependence of diffusion rates are not anticipated, although the oxygen fugacity dependence may change. For calculations in natural systems the present data would yield higher closure temperatures, longer timescales, slower cooling rates, and shorter length scales of diffusion compared to those obtained using earlier diffusion data.

No abstract