was that which would have existed at the onset of fallout at each location. Correction for decay leads us to estimate a slightly steeper rise to the curve for cumulative percent activity versus time post-BRAVO detonation. Eighteen Rongelap people went to Sifo Island, Ailingnae Atoll, to fish and to make copra (Sh57). They left Rongelap Island prior to or at about the time of the BRAVO detonation (Sh57). These people were located about 150 km (90 miles) from the detonation site at the time of fallout arrival, and thus, would have received fallout similar in granule size to that shown in Figure 4 for Bikini ash. Their location was further south than the 5th Lucky Dragon, however, and much less debris per unit area ultimately fell on Sifo Island. The fallout encountered by these 18 people was estimated to be off the centerline of maximum activity of the BRAVO cloud by about 30 km (20 miles). d. A Simple Model for Exposure-Rate History at Rongelap, Sifo, and Utirik Islands. We combined the exposure-rate survey by the radsafe team of the USS PHILIP (0C68), the early fallout decay exponents indicated by Miller (0C68), and the time of arrival of fallout at Rongelap to estimate the exposure-rate history prior to evacuation (see Table 11). This exposure-rate history would not include the contribution from non-fission-fragment nuclides since it was an extrapolation which we based on the fission product decay exponent given by Miller (0C58). The total integrated exposure at 1 m above Rongelap Island, from the time of onset of fallout until evacuation, was 4.6 x 1072 C kgm! (180 R). (This compares to 7.2 x lo-2 c kg! (280 R) based on Bikini ash results; details of this estimate are given in upcoming sections.) We estimated the mean exposure rate to be 3.1 x 1072 c kg! h-! (12 R h7!) at 5 hours post-detonation. We assumed for this estimate of peak exposure rate that all the fallout was on the ground at the onset, that is, an instantaneous rise in exposure rate from zero to its maximum value. The radsafe team of the USS PHILIP surveyed Sifo Island and the radsafe team of the USS RENSHAW surveyed Utirik Island during evacuation efforts (0C68). The exposure-rate histories for both islands were estimated from these survey results and are given in Table 11. As before, the exposure rates listed in Table 11 were based on Miller's decay estimates and we do not account here for buildup and peaking of radioactivity. Sondhaus reported that surveys at evacuation were performed with only one survey meter (AN/PDR-39), and that this instrument had not been calibrated recently and its operating condition was not known to be satisfactory at the time of use (Cr56). Sondhaus, therefore, based his estimates of external exposure on surveys made with calibrated instruments 7, 8, and 9 days after BRAVO detonation. Estimates of exposure in Table 11 which we made with evacuation survey results are, however, in agreement with estimates of cumulated exposure made by Sondhaus. Thus, we conclude the survey results by the radsafe teams were accurate. A more refined estimate of external exposure-rate history and cumulated external whole-body absorbed dose performed by us was based on Bikini ash results as follows. - 28 -