31 usually associated with a consistent eosinophilia, and indeed a large group ofindividuals with high eosinophil counts hadstools negative for parasites. However, the greater incidence of eosinophilia among Marshallese with stools positive for T. trichiura indicates that infection with this helminth may be a contributing factor, but this does not entirely explain the generally high incidence noted. Possibly chronic infections, particularly fungus infections of the skin, may be partly responsible. Anotherpossibility is trichinosis infestation, which has to be consideredseriously in view ofthe large numberofrats on the island and thepresenceof swine (used to a small extent for meat) roaming freely. On the next survey serological tests for trichinosis antigen will be carried out. An unexpected finding was that the level of serum protein-boundiodine in these people was significantly above the normal range. Butanolextractable iodines on 6 cases also showed values at the upper limit of normal, but thyroxinebinding capacity determinations on 12 cases gave data inadequate to define precisely whetherthe slight elevations weresignificantly different from normal. However,it could be calculated that the level of thyroxine-binding protein was insufficient to cause the elevation of serum thyroxine (presumably to maintain a normallevelof free thyroxine) noted in these people. The study of genetically determinedtraits has proved most interesting in helping to establish the anthropological background of the Marshallese people and the homogeneity of the population ‘understudy. Interesting findingsin the studies of blood groupings were the high frequency of the O gene (78.9%), the extremely low frequencyofthe M gene (14%), the highest incidence yet reported of the R’ chromosome (98.5%), the presence of 10.8% of Duffy (Fy*) negatives, the absence of Kell and Diego factors, and a single sample of the A.B group. These groupings mostclosely resemble those of the people of Southeast Asia and Indonesia. Haptoglobin studies showed a very high incidence of the 1 —1 type and the Hp’ gene exceeded only by that of the Yorubas of Nigeria. No unusual hemoglobin types were noted. These findings suggest a rather homogenouspopulation. RADIATION ECOLOGICAL STUDIES It seems appropriate to discuss the Marshall Island data as part of the world-wide fallout problem. There has been much concern expressed both in scientific journals and in populararticles about the hazard from fallout, particularly Sr®. The general situation as of mid-1957 has been reviewed by Robertson and Cohn,"* with the conclusion that existing levels of radiation from fallout add little to the environmental radiation hazard. Eisenbud and Harley“ present data indicating that in the United States Sr*’ continues(in 1958) to be deposited at a rate of 11 to 54 mC/mi*. The averagefor the rest of the Northern Hemisphere is 16 mC/ mi‘, which is about twice the valuefor the Southern Hemisphere. Kulp and Slakter*’ conclude thatthe diet of an average U.S. citizen in 1957 contained about 6.5 puSr°’/g calcium, which corresponds to an equilibrium base level of 1.6 puC/g if the discrimination factor between diet and boneis +. Finkel,°* in an appraisal of the potential Sr’? danger based on data from animal experiments, concludes that the minimumeflective dose in man maybe a burdenoffrom 5 to 10 uC Sr°’°, in close agreement with an estimate of 6 to 15 wC based on the radium methodof extra- polation. Hindmarshet al.*? have re-evaluated the relative hazards of Sr’° and Ra**". Their conclusion is that the currently accepted maximal permissible dose figures for Sr’ are substantially correct. Brues’* reviews the arguments upon which is based the fear that very low doses of Sr" might producea “very low (but in absolute numbers appreciable) incidence of leukemia” and concludes that the present datafail to indicatea linearrelationship for dose and effect at low doses. He further emphasizes the fact that there are other theories of the etiology of cancer, and thattheir existence weakens the arguments of those who would assign unrealistically high probabilities to the role of single mutations as being the cause of cancer. Gilliam and Walter® have studied the trends in the mortality from leukemia. In most age groups the death rate has been increasing exponentially since 1921, with doubling times of about 15 to 20 years for most age groups. The younger age groups, however, have recently shown a tendency to level off, or, in the authors’ words, since 1940 there has been “‘a distinct tendency toward a decline in the rate of increase.” This tendencyis more definite with decreasing age, and in the age group 0 to | year there has been an actual decline in the death rate from leukemia. If leukemia follows from exposure to an environmentalfactor,