! here is apparanciy a decraase in activicy in near-surface sttata due <5 migration incd lower soil Levels and soil erosion. Peak concentrations cccur several centineters below the surface, wiitle the lowest soil strata disolay an exocnential decline of radionuclide concentration similiar to natterns examined in the USA at the Hanford Laboratory and the Nevada testing grounds. Figure i5 is a map of 3ixini desianating she soil sampling. locations. Tables 9 through 16 correlate results from samples common to several locations. The data reported in these tables are average concentzations for all similar samples on an. isiand. No error is reported due to the wide range of. values enccuntered within tne values selected and due to the relatively few values available for averaging. . . Lo. . Examination of the comparison data reveals thac che ratio of averages between one island and another varies relative to the nuclide selecced for a specific sample type. value. The range of ratios does tend to converge around a single For example, if the average results for Bikini Island are used as the numerator of the cratic, and the denominator is chosen to be the results from Rongelap, Sneu and Utirix, the following ratios are observed: . islands Bikini/Rongelap 3ikini/Eneu Bikini/Utirix Ratio 4 10. 20 These ratios correspond to relacive concentration differences between 3ikini and other islands in the Marshall Islands previously reported by other laboratories.