INTRODUCTION
Energy extraction from the geothermal regime in and above buried magma-chambers requires that rocks be drillable (brittle or semi-brittle) and boreholes be stable to depths of 10 km, at temperatures to partial melting, and in a natural or manmade aqueous environment. Critical are the short-term strengths and ductilities that would govern during and immediately after drilling, and the longer-term, time-dependent response, that would influence the stability of a borehole over the multi-year life-expectancy of a geothermal well.
At least initially, we chose to investigate the short-term failure strengths and strains at failure of room-dry and water-saturated, cylindrical specimens (2 by 4 cm) of Charcoal Granodiorite (CG), Mt. Hood Andesite (MHA), Qnd Cuerbio Basalt (CB) at a strain rate of 10-4s-1, at effective confining pressures of O, 50, and 100 MPa and at temperatures to partial melting. Previous laboratory work needed to be augmented because virtually all testing above 500øC had been done at either atmospheric pressure (e.g., Murrell and Chakravarty, 1973), at too high a confining pressure (e.g., Handin, 1966; Carter, 1976; Tullis and Yund, 1977; Carter and Kirby, 1978; Tullis, 1979), or at uncertain effective pressures (e.g., Murrell and Ismail, 1976; Tullis and Yund, 1978; Van der Molen and Paterson, 1979). Available 'data on creep of crystalline rock have suggested that deformation rates are very slow even at elevated temperatures and pressures (Carter and Kirby, 1978; Handin and Carter, 1980; and Carter et al., 1981). Only when partial melting occurs do rock-strengths vanish and relatively low viscosities obtain (Van der Molen and Paterson, 1979; Friedman et al., 1980). Hence our emphasis is on the short-term effects. Previously we dealt with room-dry specimens of these three rocks and with the Newberry Rhyolite Obsidian (Friedman et al., 1979, 1980). This was necessary because "dry-out" zones may exist immediately above buried magma and a data-base for dry rocks is needed for comparison with water-saturated counterparts in order to distinguish between effective pressure and water-weakening effects. Herein we report on experiments on the water-saturated specimens.
APPARATUS AND PROCEDURES
The triaxial-compression apparatus, including the hydraulic press and internally-heated pressure cell, the starting materials, sample preparation, and plotting of true axial differential stress (MPa) against conventional strain (percent shortening) are fully described in previous publications (Friedman et al., 1979, 1980). The accuracy of measurements of differential force, shortening, external confining pressure (Pc), and internal pore pressure (PD) is of the order of ± 2 percent. Temperature (T) is known within ± 5°C, and the maximum gradient along the -- length of the specimen is 30°C at 1000°C nominal. In brief, (1) the 2 by 4-cm cylindrical specimens are vacuum-saturated with tap water and stored under water until they are jacketed in thin-walled, annealed copper tubes and emplaced within the internal furnace of the test cell; (2) a small axial force is applied to seat and seal off the specimen and the cell is purged of air and filled with the confining medium of argon; (3) the axial force, external confining pressure, and internal pore pressure are raised simultaneously to the desired effective confining pressure, Pe = Pc-PP;