This document presents an analysis of the average and unusual locations of the Earth's magnetopause and bow shock based on data from six IMP spacecraft. The average geocentric distances to the magnetopause and bow shock near the ecliptic plane are 11.0 R_E and 14.6 R_E in the solar direction, 15.1 R_E and 22.8 R_E in the dawn meridian, and 15.8 R_E and 27.6 R_E in the dusk meridian. The bow shock hyperbola is oriented in a direction consistent with the expected aberration of a radial solar wind. Observed magnetopause crossings agree well with theoretical predictions in the noon meridian plane but fall outside the theoretical boundaries in the dawn-dusk meridian planes. The solar wind momentum flux is the prime factor controlling the orbit-to-orbit changes in the boundary positions. Data suggest that the interplanetary field orientation also affects the distance to the magnetopause boundary with more earthward crossings corresponding to southward fields. Six unusual bow shock locations up to 22 R_E beyond the average position are found to be due to an enhanced standoff distance associated with a low Alfven Mach number. The possibility is raised that the solar wind may have become sub-Alfvenic on July 31, 1967. The study analyzed data from six IMP spacecraft to determine the average boundary positions and the causes for their variations with time. Best fit curves obtained to represent the average boundaries in the solar ecliptic plane are characterized by geocentric distances to the magnetopause and bow shock of 11.0 and 14.6 R_E respectively near the subsolar point. The average bow shock orientation is symmetrical about the expected incident direction of the solar wind to better than 2°. The average magnetopause orientation deduced from subsolar hemisphere measurements, on the other hand, fails to reveal a corresponding aberration effect though such an effect is clear from tail measurements in the downstream region. The width of the average magnetosheath in the solar ecliptic dawn dusk plane is almost 50% greater in the dusk hemisphere as compared to the dawn hemisphere because of asymmetry of the shock. The usual theoretical methods of calculating the shape of the magnetopause appear to be adequate in the noon-meridian plane but predict a magnetopause which is too close to the earth in the dawn-dusk meridian plane. This discrepancy is probably due to the theoretical assumption that there is no bow shock and no plasma in the magnetosphere. It is suggested that the experimentally determined value of 0.33 for the ratio of the shock standoff distance to the geocentric subsolar magnetopause distance along with relatively high values of the gasdynamic Mach number together imply that an effective value of γ between 5/3 and 2 is appropriate for the solar wind-earth interaction. AnalysisThis document presents an analysis of the average and unusual locations of the Earth's magnetopause and bow shock based on data from six IMP spacecraft. The average geocentric distances to the magnetopause and bow shock near the ecliptic plane are 11.0 R_E and 14.6 R_E in the solar direction, 15.1 R_E and 22.8 R_E in the dawn meridian, and 15.8 R_E and 27.6 R_E in the dusk meridian. The bow shock hyperbola is oriented in a direction consistent with the expected aberration of a radial solar wind. Observed magnetopause crossings agree well with theoretical predictions in the noon meridian plane but fall outside the theoretical boundaries in the dawn-dusk meridian planes. The solar wind momentum flux is the prime factor controlling the orbit-to-orbit changes in the boundary positions. Data suggest that the interplanetary field orientation also affects the distance to the magnetopause boundary with more earthward crossings corresponding to southward fields. Six unusual bow shock locations up to 22 R_E beyond the average position are found to be due to an enhanced standoff distance associated with a low Alfven Mach number. The possibility is raised that the solar wind may have become sub-Alfvenic on July 31, 1967. The study analyzed data from six IMP spacecraft to determine the average boundary positions and the causes for their variations with time. Best fit curves obtained to represent the average boundaries in the solar ecliptic plane are characterized by geocentric distances to the magnetopause and bow shock of 11.0 and 14.6 R_E respectively near the subsolar point. The average bow shock orientation is symmetrical about the expected incident direction of the solar wind to better than 2°. The average magnetopause orientation deduced from subsolar hemisphere measurements, on the other hand, fails to reveal a corresponding aberration effect though such an effect is clear from tail measurements in the downstream region. The width of the average magnetosheath in the solar ecliptic dawn dusk plane is almost 50% greater in the dusk hemisphere as compared to the dawn hemisphere because of asymmetry of the shock. The usual theoretical methods of calculating the shape of the magnetopause appear to be adequate in the noon-meridian plane but predict a magnetopause which is too close to the earth in the dawn-dusk meridian plane. This discrepancy is probably due to the theoretical assumption that there is no bow shock and no plasma in the magnetosphere. It is suggested that the experimentally determined value of 0.33 for the ratio of the shock standoff distance to the geocentric subsolar magnetopause distance along with relatively high values of the gasdynamic Mach number together imply that an effective value of γ between 5/3 and 2 is appropriate for the solar wind-earth interaction. Analysis