Lithos, 23 (1989) 225-229 225 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands Discussion Vesicles, amygdales and similar structures in fault-generated pseudotachylites - Comment JACQUELINE EABY DIXON ~ and TIMOTHY H. DIXON 2 Z Division ofGeological and Planetary Sciences, Cal~[ornia Institute of Technology, Pasadena, CA 91125 (U.S.A.) "Division ~f Earth and Space Sciences, Jet Propulsion Laboratory. Cal~/brnia Institute of Technology, Pasadena, CA 91109 (U.S. .t.) Received November 10, 1988; accepted April 7, 1989) Fault-generated pseudotachylites form by fric- tion-induced melting during earthquakes. Their study may provide important information on earth- quake processes (e.g., McKenzie and Brune, 1972; Sibson, 1975; Grocott, 1981). Maddock et al. (1987) proposed that the relative abundance of amygdales and vesicles in pseudotachylites could be used to constrain their formation depth, and esti- mated that pseudotachylites in the Ikertoq Shear Belt in Greenland formed at a depth of 1.6 km. Similarly, the absence of vesicles in the Harry Creek pseudotachylite in central Australia has been used to infer a formation depth in excess of 2 km (Allen, 1980). These depth estimates implicitly assume that bubble growth occurs under conditions of vapour- melt equilibrium, so that the volume of exsolved vapour (determined from vesicle abundance) rep- resents the concentration of volatile elements in the original rock in excess of their solubilities in the melt at the pressure of melting. However, we will show that attainment of such equilibrium is extremely unlikely during the rapid formation of pseudota- chylites. If the rate of quenching melt to glass is more rapid than that for bubble growth, the presence of large amygdales and vesicles in pseudotachylites will not be controlled by exsolution of CO2 and H~O from an oversaturated melt, and the occurrence or lack ofamygdales or vesicles cannot be used to infer formation depth. We do agree with other aspects of the model pre- sented by Maddock et al. (1987). As noted by these authors, amygdales in the studied pseudotachylites can be distinguished on mineralogical and petro- graphic grounds from other superficially similar features such as porphyroclast replacement pseu- domorphs. The occurrence ofplagioclase microlites and their textural relations relative to the amyg- dales provide good evidence that pseudotachylites were mainly liquid at some time, and that amyg- dales formed as bubbles in that liquid. The liquid is assumed to be generated instantaneously during earthquake faulting, and undergoes a volume in- crease relative to the original solid. Liquid pressure briefly exceeds lithostatic pressure, leading to ten- sile failure of adjacent rock and formation of injec- tion veins. Liquid pressure then falls to the lithos- tatic value. It is at this point, Maddock et al. ( 1987 ) suggest, that vesiculation occurs. Freezing of the in- jection vein preserves the vesicles. Later mineral growth forms amygdales, but in general the size and shape of the original vesicles are retained barring later tectonism, enabling volume estimates of the original volatile phase in excess of the solubility at that pressure to be made. The mass fraction of vol- atile components in the original melt is estimated from analyses of the country rock. The exsolved va- pour component, estimated from the volume of amygdales, is subtracted from the initial volatile content of the melt to obtain the final volatile con- centration. This concentration is compared to pres- sure-dependent solubility curves to obtain the equi- librium pressure. We agree with this model until the point where injection veins form and liquid pressure falls to lithostatic values. We doubt that significant vesicu- lation can occur at this point, regardless of the de- crease in liquid pressure, because of the kinetics of bubble growth. Whether significant vesiculation can