This study was conducted in order to characterize the frequency, orientation, and aperture of macroscopic fractures in the crust and their effect on physical properties over an appreciable depth interval (1829--3450 m). The following are our major findings: (1) Over the range of apparent apertures measured with confidence, the frequency of fractures with a given aperture decreases as aperture increases. With applied corrections for sampling bias, the observed distribution of fracture aperture has a power law form providing evidence of the self-similar nature of fractures in the crystalline crust. Fractal analysis of the fracture aperture data yields a fractal dimension of 1.4 over the range of reliable aperture measurements in this study from 15 to 100 mm. (2) Fracture frequency does not systematically decrease with depth in the study interval. (3) No significant correlation was found between fracture occurrence and lithology, and both fracture spacing and aperture are uncorrelated with fracture orientation or depth. (4) The majority of fractures encountered in the well strike NNW-SSE and dip steeply to the west. One set of steeply dipping fractures appears to be related to the NW striking San Andreas fault and appears to be related to steeply dipping, NW striking shear fractures observed in nearby outcrops that are characterized by laumontitic alteration. (5) The fractures bear no obvious relation to the current northeast direction of maximum horizontal compression but do correlate with anomalies in physical properties measurements of compressional and shear velocity, porosity, and resistivity. (6) Macroscopic fractures strike in a direction nearly orthogonal to the fast propagation direction of seismic wave anisotropy determined from vertical seismic profiling experiments in the well. These fractures appear to be unrelated to the observed seismic anisotropy. (7) Hydraulically conductive fractures and major faults indicate that fluid-conducting fractures are a subset of the overall statistically significant population and not related to the San Andreas fault or to the orientation SHmax in an obvious way. ¿ American Geophysical Union 1992