underground instrumentation, and data from Station 650.05 on Aomon are lacking because of recorder failure. . This study is limited, consequently, to information from two stations, 650.02 and 650.03, which should be pertinent, and a third, 603, of doubtful pertinence. The latter data, although highly significant to the reaction of surface structures of the type used for recorder shelters at Pacific Proving Grounds, can be related to ground motion in this study only by crude extrapolation from shorter ground ranges. Results from the most remote instrumentation, that on Parry, could at best have been of academic interest and, because of a too optimistic estimate of set range, were burdened by low signal-to-noise ratio and serve principally as tests of integration techniques (Appendix B). Ground motion of sufficient magnitude to damage underground structures in the Pacific Proving Grounds evidently did not occur at scaled ground ranges (R/W" in ft/lb”) as great as 6.6 (Engebi):where the recorder shelters were intact following Mike shot. However, the earthcovered recorder shelter on Bogon, Station 600, at a scaled ground rangeof 3, was subjected to such severe motion, probably from incidence of the air shock, that failures were produced in the recording circuitry and stee] doors were jammed. Correlation of ground-motion data with damageat these stations is of doubtful significance because of the incompleteness of information from the instrumentation at Station 650.01 near the recorder shelter on Bogon. 1.7 CONCLUSIONS 1. Data from only two of the six stations instrumented were wholly suited to the purpose of the ground-motion study. 2. Information from these two stations and such partially usable data as were available from other stations suggest that, for the Pacific Proving Grounds, maximum ground-transmitted accelerations from weapons burst above ground may be estimated from the empirical equation Aw? = Kap! where acceleration is in g units and the scaled ground range A is derived from w, the cube root of the radiochemical yield expressed as pounds of TNT. The coefficient K has values be- tween 2.66 x 10° and 2.1 x 10° for scaled heights of burst between 0 and 0.14 ft/lb”, 3. Acceleration induced in the ground at Pacific Proving Grounds by locally incident air shock is related to the peak air overpressure by the equation A= 0.053p'-? for accelerations in g units and pressure in psi, with an error of +30 per cent. 4, Motion of elements of a massive, rigid structure founded on loose water-filled sands may be from 2 to 6 times greater than the motion of the soil, although the frequency of the motion will be similar to the latter, according to data from Station 603. Motion of such structural elements induced directly by air shock appears to be considerably greater than ground motion from the ‘same source and will include a frequency characteristic of the structural element. REFERENCES 1. P. A. Northrop, Instrumentation for Structures Program, Operation Greenhouse Report WT-1, January 1951. 2. H. E. Lenander, R. S. Millican, and D. E. Showalter, Instrumentation for Blast Measurements by Sandia Corporation, Operation Ivy Report WT-606, December 1952. 3. J. J. O’Connor and D. R. Powers, Lt USAR, Ground Shock Measurements, Operation Greenhouse Report WT-69, August 1951. 28