$.1.$ Combined Cloud and Fallout Data. ¥ alternative Processes to fallout are not import fission products with volatile predecessors can be os useful as gaseous fission products for measuring the extent cf fallout. Because it ta incorrect to assume that the ne content of a volatile fission product in fallout is zero, the R-value in faliout must be measured Then: y= RYeg - (RU) £0 . [R* (Ye — fre (Y)) FO This formula can be derived by algebraic operations from the definitions of the R-values . (Appendix E). If, despite the fact that it is incorrect, the R-value for Y tn fallout is assumed to be zero, the above equation reduces to the expression for a gas, and y becomes the upper limiting value for the fraction of Mo”(or refractory debris) left in the region sampled Fission products such as Sr™, Ca¥’, and to a somewhat lesser extent Sr® appear to behave very muchlike Kr" in Shots Koa, Walnut, and Oak and may be used to estimate fractional fall. out of refractory debris or upper limits to the fraction remaining aloft. The disadvantage of using Sr"? or Cs! for this purpose Is that R-values must be measured in fallout and are necessarily constant. The chief advantage ia that the analyses may be extended to longer times, because the half-lives are long and a sufficient sample may be obtained by almply filtering more air. Values have been calculated in the above manner and are given in Table 3.7. In calculating the values for fraction of Mo™ in the cloud, the data must be picked from Tables B.1 through B.6 with care. Only cloud samples taken in the light and variable layers are used, and these are matchedon an indivudual basis with height line samples taken at a later time, wherever possible. The half-lives of the noble-gas precursors of the nuclides used above are: Cs'*’, 3.8 minutes; Sr**, 3.2 minutes; Sr™, 33 seconds; Y"!, 10 seconds; Ce“, ~1 second; Cs’, none. The frac- tion of Mo™ remaining in the cloud as calculated by each of these nuclides generally increases inversely as the half-life of the nuciide’s noble-gas precursor. [If it is assumed that the Rvalues in the height line samples are representative of the material that has fallen from the light and variable layer, the results of the calculation of the fraction of Mo" remaining in the cloud may be interpreted to mean that the original R-values in the light and variable layer were not representative of the device. This is due to the fact that if the original R-values were representative and if the average R-value is used for all the fallout, the fraction of Mo’ calcu- lated to remain in the cloud (y) should be the same no matter which radionuclide is used in the calculation. However, the same experimental data could have been obtained if the eampled region origin- ally had representative R-values, provided the R-values from the height line samples were not representative of all the fallout from the light and variable layer. The assumption here is that the unsampled portion of the fallout, i.@., the portion between 1,000 and 50,000 feet, had Rvalues between those found in the fallout and in the cloud. The explanation of such behavior might be that nuclides that condense shortly after the explosion occur in larger particles than nuclides that condense later, e.g., those with noble-gas precursors. The larger particles fall faster, are depleted in the cloud samples, and are enriched in theheight line samples. The opposite situation would exist for small particles. The actual explanation of the variation in the calculated fraction of Mo" remaining in the cloud may well be 2 combination of the two given above. Smail variations, such as those due to experimenta. uncertainties in the R-values, have large effects on the calculation when the differences between the device R-values and those observed in the cloud and fallout are small. The Mo’fractions calculated from Cs‘’" and Sr, the two nuclides having the longest-lived noble-gas precursors and showing the greatest fractionation, are given in Table 3.8. They are compared to the Mo” fractiona calculated from Kr®?, 39