with the cloud for longer periods of time and possibly absorbed more longerlived nuclides than did the very large granules. In the forthcoming analysis, the fractionation observed for Bikini Ash granules was assumed to be similar for granules at Rongelap, Sifo and Utirik Islands. With the possible exception of Utirik Island, this was considered a value assumption due to the proximity of Rongelap and Sifo Islands to the 5th Lucky Dragon. 2. The Decay of Fallout The gamma and beta decay of the BRAVO radioactivity after the first 10 days post detonation was measured by many researchers (e.g. Miller, Servis, Tomkins, Wilsey and Stetson, see 0C68).- Decay data measured prior to 10 days was not found in the literature. Fallout samples, taken weeks after the BRAVO event, were from Bikini Atoll, Rongelap Atoll and from the surface of US Navy ships in the area. The measured decay exponent after two weeks was used by many researchers to extrapolate exposure rate back to times prior to sample col-~ lection and in one case was used to estimate activity decline every hour post detonation (Miller (0C68)). These calculations by Miller for the decay of fall- out activity from several hours out to a few weeks post BRAVO apparently excluded the decay characteristics of non-fission nuclides. This would impact on surface activity estimates at the islands of interest since these estimates relied on extrapolated exposure rates. The thyroid dose from internal sources would be affected also by decay characteristics because it relied in part on sur~ face activity estimates. In order to derive ground activity estimates at times close to BRAVO detonation and to derive external and internal thyroid dose, the gamma or beta decay rate decline over short periods of time was assumed to have followed the relationship