left calf circumference, biacromial diameter, and bicristal diameter. Standardized techniques (47-49) were used. The status of secondary sex characteristics was evaluated by inspection with the methodology described by Greulich et al. (50,51), Reynolds and Wines (52,53), and Shuttleworth (54). History of menarche in girls, hair distribution, breast development, and penile and testicular development in boys were recorded during these examinations. In 1958 apparent discrepancies regarding birth dates were noted in the charts of many children. The absence of recorded birth information in the Marshall Islands seriously complicated the verification of ages. Detailed genealogical and biological histories were compiled for the Rongelap population in 1958-1960. The reconstruction of birth chronologies was based on intensive evaluation of frequently contradictory information derived from the following sources: * Dates of birth reported by parents. * Dates of birth recorded occasionally in ledgers kept by the village magistrate. * Limited number of birth certificates (not always accurate) on file at the courthouse in Majuro. * Birth order of children within each family unit. * Ranking of childhood population in terms of age by parents. * Ranking of childhood population in terms of age by the children (particularly age peers), by relatives, and by friends living in the village. A table of most probable birth dates was derived for the Rongelap childhood population. Biologic compatibility of the birth dates within each family was carefully checked, and the compatibility of physiologic status with age was also determined for each child. The presumptive dates of birth, shown for each individual in Appendices 11, 12, and 13 of ref. 11, have been used in the calculations of chronological ages for analyses of growth and development. Roentgenographic documentation of osseous maturation (x raying of the left hand and wrist) was initiated in the exposed children in 1957. A major effort was made in 1958 to examine the skeletal maturation cf both exposed and unexposed children. Unfortunately these valuable baseline films were lost at sea during transport. This created a gap of almost 3 years when no radio- graphs were available on a number of children in the spurt phase of growth. Thereafter roentgenographic studies of the left hand and wrist were included at irregular intervals. These were particularly difficult under field conditions and presented many technical problems, but the minimal number of roentgenograms eventually obtained permitted a reasonable assessment of the longitudinal skeletal development of each child through the chronological age of 16 or 17 years in the girls and 18 in the boys. Skeletal age determinations were made by inspection with the techniques and standards published by Greulich and Pyle (51). Early analyses of the skeletal age data were included in the reports of previous surveys (11-13). Comparisons between the exposed and unexposed children were made primarily in the group who were <10 years old on March 1, 1954. This group was further sub- Lo Ce Cu: cm cD divided into two categories: those <5 years old on March 1, 1954, and those aged 5 to 10. The data on children >10 years old on March 1, 1954, could not be analyzed in detail. The number of children (particularly in the exposed group) in each age category was extremely small. Also, by the time satisfactory -27-