fallout from food was reported by persons evacuated from Rongelap or Sifo Is- lands (Sh57). Fallout was distributed on the surface of Rongelap Island at 12 hours post detonation at a level of about 1 Ci m2 (see Table 5). This was in good agreement with soil sample data obtained on March 8, 1954 (0C68). Based on a conversion factor given by the persons doing the soil analysis (0C68) the measured soil specific activity was converted to activity per unit area at 12 hours post detonation for comparison purposes. Their value for Rongelap Island was 0.53 + 0.72 Ci m2 and was based on four samples. Considering the variables involved with the Bikini Ash estimate of activity per unit area and the variability in soil sampling (see Section III), these two estimates were in very good agreement. Random soil sampling was done at Utirik Island, Sifo Island, Eniwetak Island and other islands of Rongelap, Ailinginae, Utirik and Rongerik Atolls as well (0C68). At the end of fallout deposition at Utirik Island, estimated to be 36 hours post detonation, the surface activity based on one sample was 0.058 Ci m-, This one data point was abcut 5 times too high based on exposure rate data, however, soil analysis data exhibited wide variations in soil taken from nearly the same spot (0C68). passed by at 8 hours post detonation. At Sifo Island, the fallout cloud The activity per unit area at Sifo was measured also with one soil sample and was 0.032 Ci a2, At Eniwetak Island the BRAVO cloud was estimated to pass by at 16 hours post detonation and the measured surface activity at that time was 0.32 £0.21 Ci nm, based on 2 samples. The outside area used to prepare food for the mid-day or evening meals at Rongelap may have been about 1-2 m for a family (see Figs. 9-15). Cooking was done over an open fire fueled by coconut shells (Na80). 28 Boiling and