subsequent decrease to 7 mR/hr at D+1. Similar results were obtained by Project 2.7 (Reference 9) in the open ocean. Rapid shot-day mixing progressed in two days downto the thermocline, where the stable stratification minimized furthervertical diffusion during CASTLE. In the lagoon, contamination at the surface was observedto drift slowly westward under the action of the tradewinds. The radioactivity either adhered to the western reef, flowed overit into the open ocean,or recirculated at depth in the lagoon. There is no clear indication that the latter phenomenonled to a meaningful reappearance of contamination in the Nan area. After Shot ROMEO, which among CASTLEshots was uniquely lacking in widespread fallout in Bikini Lagoon, no reports of fresh contamination in the anchorages have been found in CASTLE documents; after other shots, reported intensity buildups are explicable by local fallout in the water that led to progressive ship contamination. The one circumstance that could have replenished the westward-drifung surface contamination was an influx from the ocean. The east-westradiation isopleths for Shots UNION and YANKEE (Reference 2) suggest this possibility; however, it would have been most pronounced for BRAVO, whereintensities increased eastward of Nan for some 100 miles. The available lagoon data thatlikely reflect this process are the 0.1 to 0.3 mR/hr waterintensities that were typically present at the Nan anchorage during CASTLE (Reference 8). Without replenishment, lagoon drift would have led to lower levels within the eights weeks between Shots BRAVOand UNION. In the mean,the reported levels are roughly consistent with decreased intensity from decay alone. 2.4 BIKINI LAGOON WATER INTENSITIES. The foregoing phenomenology and the paucity of radiological data suggest that the best available model for time dependence of water intensities is to assume no net transport of contaminants and to diminish the intensities by decay alone. This approach is most applicable for the anchorage areas and after Shot BRAVO;it likely high-sides the intensities after other shots. For the northern operating areas near surface zeros, where drift is of clearer significance in the long term, most exposures were soon enough after the shotsso thatlittle drift had occurred. Owing to the complexity of the model equations, the determination of radiation intensities from ship contamination and watershine is accomplished by numerical techniques. All logged ship movements and reported or derived water intensities are tracked throughout the 15