and then eastward within a very short time. The lowest levels continued westward and the 18,000-foot trajectory appeared to curve back towards the United States on the 28th, although the meteoro- logical data is uncertain. Winds in the stratosphere up to the level of the ton of the cloud were probably from the east, carrying most of the mushroom westward, Although almost a month elapsed between the first and second bursts of the Castle series, enoughdebris from the first burst was present to seriously interfere with attempts to trace the progress of the. second clow by an examination of fallout data. For example, an increase in deposited activity occurred at some stations in Central and South America on March 31 and April 1, several days before the meteorological trajectories would indicate the arrival of debris. It is not certain if this is due to the complete lack of meteorological observations in the Eastern Pacific and the winds were really stronger than assumed, or that the debris was actually from the Bravo burst. (Note: Since all fallout data is extrapolated to 100 days after the assigned burst, values assigned to different bursts cannot be compared directly. The extracclation factor denends both on the day of the burst and on the day the sample was counted. For the areas mentioned in this saragrarh, values assigned to burst 2 would have to be increased by about a factor of three if the debris were assignedtabursti). By April 2nd and 3rd increases in activity are evident along the ' Gulf Coast of the United States and certainly by the lth and 5th there is good evidence that debris from this burst has arrived over the United States. Again, as when fresh Bravo debris was present,fallout seemedto occur irrespective ofithe occurrence of precipitation. The progression of debris westward from the. test. site appears to. have been more rapid than indicated by the low-level trajectories at 5,000 and 10,000 feet, at least for the first few days following the burst. Whether the arrival of debris at Yap and Koror on March 29 is a result of transnort of material westward in the stratoscheric easterlies or in faster-than-observed low-level trades is not certain. Again, as with Bravo debris, there was 2 marked tendency for the fallout to occur in the tropical areas, with occasional incursions into the United States. 2.3 KOON Koon, the third burst of the series ; . was detonated at Bikini 31820 GCT, Avril 6, 19S). but cloudy conditions prevented - ll - SS