operational system, a rough estimate of yield within about an order of magnitude may be obtained from broad-band field-strength measurements with proper correction for path, terrain, ionospheric conditions, time of day, etc. However, the corrections to be made were imperfectly known. Frequency analysis of wave forms. together with other characteristics, may offer some assistance. Field strengths were measured at various places, but variations with presumably identical equipment at the different iocations were not all explainable. There appeared to be un approximate relationship between yield and the frequency at which peak energy occurs, with some theoretical justification for this relationship. 7.1.6 Ionosphere Data. layer height of about $0 km The arrival times of the first sky wave gave an ionospheric Some records showed as many as five sky waves, but of course with less energy for cach reflection; they also indicated a layer height of about 80 km. 7.1.7 Peripheral Lightning. Fast-trame moving-picture photography (3,000 or more frames per second) ofIvy Mike had shown what appeared to be lightning flashes between the natural cloud cover and the sea cn the periphery of the fireball. This phencmenon Started at about 5 msec after the beginning of the nuclear reaction and continued for about 75 msec or more. These visible flashes were also in evidence on Castle high-snced photographic film. No signals attributable to the discharges were noted. 7.2 AIRBORNE LOW-FREQUENCY SOUND Acoustic measurements from remote stations had been made, prior to Castle, on ail nuclear tests except Trinity. The purpose of the experiments carried out during Crossroads, Sandstone, and Greenhouse had been to establish the feasibility of detecting nuclear explosions of moderate yield at ranges in excess of 4,000 km by acoustic means —felt to be the minimum range at which a suitable acoustic system for detecting foreign explosions could be established. Results from Crossroads and Sandstone had indicated positive detection to a range of only 1,900 km. With improved equipment and better techniques, detection had been accomplished out to 4,500 km during Greenhouse. Additional experiments had been carried out during Buster-Jangle, Tumbler-Snapper, and Upshot~-Knothole to delineate the capabilities and limitations of acoustic-detection techniques for a wide range of yields of air, surface, and shallow-underground detonations during different seasons of the year. Results from these tests indicated a limited, but usable, detection range for low-yield explosions—— even for shallow underground detonations. Seasonal shift in propagation, which had originally been noted during tests conducted with small TNT charges, were confirmed. It had been found that amplitudes varied considerably with propagation conditions and that any correlation between signal period and yield was quite variable. Results from experiments carried out on Ivy had indicated that acoustic signals from high-yield kiloton and megaton explosions were detectable at longer ranges and showed 94