collected, read, and then ret was conducted for five weeks, (lowest curve) are the mean outdoor natural y-levels obtained from the Health and Safety Laboratory spectrometer readings. These values are directly proportional to mean soil content of natural radioactivity (1 mr./week -> 7-83 p.p.m. eV). »xposure from the mean values data in the various areas are Table 2. These air doso values ital)? by assuming that each of 1e body surface, read 100 per zation and 85 per cent of the r. The latter figure is based surements of body attenuation of population exposure data show a considerable sys- z of 400 individuals, 25 in each each of the eight geological this study are discussed by on’, PAM y Laboratory and Harvard s are plotted as a function of dioactivity in Fig. 1, with the idicated. Plotted also in Fig. 1 J 30 40 50 60 70 < radioactivity (p.p.m. eU) in the eight selected areas of northern itimated mean bedrock radioactivity. } trom Harvard dosimeters; B, popu| Health and Safety Laboratory in situ in weekly outdoor y doses 6 It is readily apparent from tho figure that the two sets tematic differonce. The large intercept at zero bedrock radioactivity for the linear fit to the Harvard dosimeter data cannot be explained simply on the basis of the relatively constant cosmic-ray and fall-out dose contribu- tions, which are less than 1-0 mr./week. Thereis certainly no evidence to suggest that building materials produce elevated radiation levels indoors in any consistent manner. The intercept for the Health and Safety Laboratory results is more reasonable, although also somewhat high. It is interesting to note that, while an apparent linear trend appears to exist for all sets of data, the interpretation of this trend is not obvious. For example, the slope of the natural y-dose regression line is only one-sixth of that expected if the bedrock were the source of the radiation". In a sense, the results shown in Fig. 1 provide a rough indication of the influence of bedrock geology on soil radioactivity and natural radiation exposure in these areas. The effect is small and may beof practical significance only in the Conway area, where the reddish sand derived from the thorium-rich Conway granite is present in the soil throughout the populated areas. The trend of the results in Fig. 1 may be indicative of some fairly consistent relation in these areas between the bedrock formations and their respective overburdens of soil in terms of natural radioactivity. The apparent nearlinearity of the population exposure estimates as a function of bedrock radioactivity derives from the similar relationship between estimated mean soil and bedrock radioactivities, since the outdoor (and to some extent the indoor) radiation-levels to which the general population is exposed are closely related to the content of natural y-emitting radioisotopes in the upper layers of the soil. Fig. 2 shows the Harvard dosimeter data plotted directly as a function of the Health and Safety Laboratory 1962 total exposure results. The high degree of correlation (r > 0-9) between the Harvard and Health and Safety Laboratory estimates of population exposure is evident; a line of unit slope fits the data quite well. The 1-2 mr./ week value for the Y-axis intercept of this line is a measure of the apparently systematic deviation between these two sets of data. While not enough information is at present available to explore this problem fully, one obvious possibility is that the pocket dosimeters consistently exhibited enhanced leakage undor field conditions as compared with that measured in the laboratory and corrected for in the data interpretation.