useful only to indicate order of magnitude values of 5556 soil : concentrations present at the collection time. Unfortunately it is not possible to offer a clear argument in explanation of the > Fe body burdens of the Rongelapese presented here at this time. Samples from 1963 through 1969 would have shed Light on the problem, but none are available for analysis, the possibility of Rongelap lagoon acting as a nutrient and trace-element "trap" similar to estuaries intriguing. (17) “is Removal and retention of both stable Fe and S56 ‘from the North Equatorial current could lead to high specific activities of the radionuclide in species important in the - Rongelapese diet. The fact that livers from mature goatfish contain between 2-3 times as much stable iron per unit wet weight as do livers of mature salmon (4) is consistent with this argument. Marshallese in general show a tendency toward anemia (9) and thus may absorb more iron from their diet than do non-anemic individuals. The Phe body burdens of the Rongelapese May there- fore only reflect more complete uptake of iron rather than uptake _ of iron of high specific activity. It is clear that further | Measurements of the specific activities of S56 in the diets of _ the Rongelapese and the effective half-life they display for this radionuclide will be needed to clarify these possibilities. A tnene me