Journal of the Geological Society, London, Vol. 157, 2000, pp. 1093–1096. Printed in Great Britain. Deformation mechanisms and rheology: why marble is weaker than quartzite K. H. BRODIE 1 & E. H. RUTTER 1 1 Department of Earth Sciences, University of Manchester, Manchester M13 9PL, UK (e-mail: kbrodie@fs1.ge.man.ac.uk) W hen deformed together in nature, calcite rocks invariably appear weaker and more ductile than quartz rocks. This can be reconciled with exper- imental flow data only by taking into account grain-size-sensitive (GSS) flow in calcite rocks, which is pre- dicted to dominate even at grain sizes on the order of 1 mm at middle metamorphic grades. Using new experimental data that demonstrate the transition between intracrystalline plastic and GSS flow of quartz rocks, we predict that unnaturally small grain sizes at temperatures of 700C or higher are required for GSS flow of quartz in nature. Thus natural flow of quartzite is expected to occur by intracrystalline plastic processes, even after recrystallization to a fine grain size. Keywords: deformation, marble, quartzite, plastic flow, grain-size- sensitive flow. In many studies of naturally deformed quartz and calcite rocks, microstructures have been interpreted in terms of defor- mation mechanisms. Evidence for intracrystalline plasticity (crystallographic and grain-shape fabric, subgrains and other intracrystalline strain features, ‘creep’ dislocation structures at the transmission electron microscope (TEM) scale, etc.) may be unambiguous, although some can be removed by post- tectonic annealing. Evidence for grain-size-sensitive flow in the form of diffusion- or dislocation-accommodated grain- boundary sliding includes fine-grained, equigranular micro- structure in a rock displaying macrostructural evidence of high strain (extremely attenuated folds, shear zones), lack of pre- ferred crystallographic orientation (CPO), evidence of cavi- tation at grain boundaries (Behrmann & Mainprice 1987), or particular distributions of crystallographic misorientation between grains (Jiang et al. 2000). These mechanisms are not mutually exclusive, and their contributions to total strain rate are additive, although one or the other may dominate at particular physical conditions. Some fine-grained rocks may display equigranular microstructure accompanied by strong CPO, or TEM evidence of a heirarchy of different levels of development of dislocation density in different grains, that can be interpreted in terms of cyclic plastic working and dynamic recrystallization by grain boundary migration. Where they occur in proximity, calcite rocks generally appear more deformable (weaker) and more ductile than quartz rocks. Field evidence for high strain commonly goes hand in hand with evidence of tectonic grain size reduction, often to grain sizes in the range 5–20 μm. Both rock types often show evidence of grain size reduction in high strain environ- ments, but calcite marbles deformed at higher temperatures (e.g. amphibolite facies or higher) can be coarse grained and equigranular, yet show macrostructural evidence of high strain. With the aid of existing data on calcite rocks and new data on synthetic quartzite, deformed under conditions that illus- trate the transition between intracrystalline plastic flow and grain size sensitive flow, we explore the extent to which experimental studies are compatible with observations on naturally deformed rocks. Previous data on intracrystalline plasticity. A number of experi- mentally determined flow laws for intracrystalline plastic flow (power law creep) of calcite and quartz rocks have been published. The range of predicted flow stresses at a ‘natural’ strain rate of 310 14 s 1 for quartz is enormous (see Paterson & Luan (1990) for summary). Figure 1 shows the range of these predictions as a shaded area. All are character- ized by a stress exponent of 4 or less. The more recent studies are arguably the more reliable, and these tend to fall at the upper bound to the shaded area (Luan & Paterson 1992; Gleason & Tullis 1995). Post et al. (1996) showed that quartz rheology is sensitive to water fugacity, but the above extrapo- lations do not apply to identical values of water fugacity. Paterson (1986) argued on theoretical grounds that the equi- librium solubility of water in quartz even at high pressure and temperature will always be less than 100 H/10 6 Si. Hence it is presently difficult to extrapolate with confidence an empirical flow law in which deformation rate is proportional to water fugacity to some power, when the uptake of large, metastable Fig. 1. Comparative flow behaviour of quartz (broken lines) and calcite rocks (solid lines) deforming by intracrystalline plasticity. Laboratory flow laws listed in Table 1 are extrapolated to a ‘natural’ strain rate of 310 14 s 1 . Power-law creep is illustrated for Carrara Marble (Schmid et al. 1980) and Yule marble (compression normal to foliation, Heard & Raleigh 1972). High stress behaviour (exponential creep and twinning, Rutter 1974) is shown for Carrara marble. Grey area contains extrapolations using all previously published quartz flow laws (summarized in Rutter & Brodie 1992). The four most recent curves for quartz rocks (plotted, after Table 1) lie towards the upper bound of the greyed area. Extrapolations of the marble data overlap with the quartz curves, inconsistent with geological observations. 1093