dependence was 1.01 to 1.05. The mean correction was approximately 1.02. Conse- quentiv, we felt no need to correct the remaining 1976 or 1977 data for the minor energy dependence encountered. Results A total of 112 RSS-111 measurements were taken on five atolls. point is the average of at least 20 individual readings. Each data This assures the precision of the value while the initial calibration guarantees accuracy. one sigma error is on the mean exposure rate. All exposure rate values include nacural background except where otherwise noted. Figure 1 graphically presents the data obtained at Eniwetak Island, Rongerik Atoll. On this island, random measurements were taken along a central northsouth transect. the raw data collecced with one sigma error. island is 6,3 +R/hr. The Table | presents The average exposure rate for this This is abouc 1.5 times higher than the cosmic/terrestrial dace rate found on uncontaminated coral islands, in the Rongerik Atoll due to presence or U, S$. Eniwetak was the island surveyed servicemen at the weather station there at the time of the BRAVO fallout incident. Tables 2, 3, 4, 5 and 6 present the raw data from Rongelap Atoll. surveyed were Kabelle, Naen, Eniaetok and Rongelap. The islands Naen is located at the northwest corner of the atoll, and Kabelle at the northeast corner. Kabelle is a significant copra resource; and both of these islands may be used for brief visits, but neither of them is permanentiv inhabited. These islands received a Significant amount of fallout debris and consequently, are still substantially more contaminated than the islands of Rongelap and Eniaetok, aast and eastern parts of the atoll. The current values for external exposure rates on these islands are listed below and in Table 14. Lsland located in the south- The entire population presently Average Exposure Rate in -R/hr Naen 43.1 Kabelle 21.7 Eniaetok 9.9 Rongelap 7.3