any previous nuclear explosion, any attenuation of pressure (as compared with that predicted) or any appreciable deviation of the observed pressure-time-distance curve from the “ideal” could passibly be attributed to phenomena arising out of the size of the explosion. Effects of ambient pressure and temperature gradients, such as are characteristic of the upper atmosphereof the earth, upon a passing shock wave are virtually unknown. After the shock wave has degenerated into a sound wave, its path may of course be calculated. Any ex- istent theories regarding the effects of these temperature and pressure gradients on shock waves have as yet to be substantiated by reliable experimental observations. Available infor- mation does make possible some qualitative deductions” !* about the possible distortion of a spherical shock wave in a nonhomogeneous atmosphere which might explain a decrease in peak overpressure as measured at ground level, an attenuation which could conceivably become increasingly effective at greater distances from ground zero. Since it is known that water is nearly an ideal reflecting surface, it was predicted that overpressures measured on the blast line over water on King shot would be nearer the ideal or estimated overpressures than those measured on the land blast line. Not only does water re- flect a much greater percentage of the thermal energy incident uponit, but, in addition, any thermal energy absorbed by water is absorbed not just in the first fraction of an inch at the surface but over the entire path length of the refracted ray. As contrasted with observed behavior over land blast lines, one would expect to find no evidence, on bursts over a water sur- face, of precursor formation,'.?."" the thermal-mechanical effect in which the surfacelayer of soil is heated to the point of exploding or forming clouds of dust which movein the path of the shock wave. It would be virtually impossible to heat the huge mass of water involved to the vaporizing point under the conditions of this test. $ Strictly speaking, the land blast line used on King shot differed somewhat in configuration from those used on similar tests at the Nevada Proving Grounds. Because it was necessary to place the measuring stations at different azimuths from the reference line passing through ground zero and because of the shape of the island on which stations were placed, some water was interspersed among the land areas; thus the paths traversed by the shock wave in reaching the various stations had different ratios of land to water. Nevertheless it was believed that conditions were similar enough to those in Nevada to warrant valid conclusions regarding the effects of different types of terrain on blast-wave propagation. It was hoped that comparison of measurements on the land blast line with those made at corresponding distances over water would give some indication of the degree to which pressures were attenuated as the shock wave passed over land areas and would serve as a semiquantitative measureof the deterioration of the pressure-time curves from the ideal waveforms as a result of thermal and terrain effects. i Inasmuch as the scaled height of burst on King shot was low —178 ft—the results of these measurements are not particularly significant so far as supplying additional points for the experimental height-of-burst chart is concerned. Actually it would have been preferable to have made this burst at a greater height, i.e., a scaled height nearer the “knees” of the height-of- burst chart, had data on height of burst been a primary objective. Requirements for other measurements, however, necessitated the limitation on burst height. 3 PREDICTION.OF OVERPRESSURES The unprecedented size of the burst and the inherent uncertainty in predicting the yield posed some new problems in estimating blast overpressures and associated thermal effects from Mike shot. Since it was to be a surface burst over water, it was felt that a reflection factor of 2 could safely be assumed,i.e., that the blast wave would take the form of a hemisphere having peak pressures, waveforms, and radii equivalent to those of a yield of twice the size in free air. The set ranges to be used for the pressure gauges, and consequently their locations, were directly dependent upon the anticipated overpressures; so it was necessary to formulate a 12 a