distance of approximately 26 miles and arrived as though generated close to ground sero, having travelled across the lagoon at the velocity of shallow water waves. Since Ivy Mike was an island shot, it was not wholly surprising that it did not generate waves in a manner ana)ogous to high explosives detonated on water. Although the Mike shot did reach into the iagoon, the generation and collapse of the cavity was not considered to be identical to that from a burst on water. Therefore, it was believed that the shot environment can~ celled out most of the direct generation region. In contrast to the Mike results, Castle data indicated that the recorded waves did ema- nate from the central region of the detonation. The first arrival was a short-period, highly damped series of ground- or water-transmitted shocks. Following these, the Overpressure , psi 2 &! Free Air from WT- 710 300 1000 Horizontal Ronge , Feet 2000 (Scoled to | kt) Figure 2.11 Averaged pressure-distance data. records clearly showed the arrival of the air-transmitted shock wave. Next, preceding the direct water wave, a slow rise in pressure (water) occurred that was postulated to be caused by large quantities of water and coral debris falling back to the water surface. This was abruptly lost in the arrival of the direct water wave —the first arrival in all cases being a crest followed by a trough. These appeared to act as oscillatory waves, the time of arrival of the first crest showing a propagation velocity fitting the relation V = gh)'/2, where h is an average depth of 170 feet assumed for the Bikini Lagoon. Refraction and reflection against a reef or shore line may significantly reduce or am- plify the destructive capabilities of water waves at termination. At Bikini, How Island is an example of a protected shore, while Nan is an example of one highly susceptible to 39