UNCLASSIFIED Early arriving particles appear to have a much higher ratio of sodium chloride to activity than do later arrivals (Table 1, last column.) Such a variation might be indicative of early large droplet fallout from the cloud stem region where the concentration of activity may be less. The sampling is small, however, so caution must be exercised in the inter- pretation of these data. The calculations of total mean values of specific activity (Table 1) do not include these initial values, nor does the calcu=lation of this value for Navajo include the approximate values defined by footnote (c) of the table, A comparison of event Flathead with Navajo shows the ratio of total yields is 13:12 while the activities produced by the devices are approximately equal, Since the barge complex for each event was identical, the insoluble solids contributed by this complex are identical for both events. On the basis of activity per gram of sodium chloride, 13 times the amount of seawater was carried aloft by Navajo. Estimates of insoluble solids specific activity indicate that Navajo fallout contained about ten times more solids per activity than did Flathead. Whether this was contributed by calcium and magnesium salts from the seawater carried up by Navajo or by additional bottom material is conjectural, It appears that the hygroscopic slurry particles can change markedly in size, density, and falling rate due to environmental influences. A detailed study of these effects is required before particle point of origin estimates can be made using measured size data, Validity of the Data The data of Table 1 are based on analyses which have been extensively calibrated and tested in the laboratory, analysis is about + 5 percent. The average error in the chloride The standard deviation error of a water volume measurement is about + 25 percent. Estimates of insoluble solids volumes are only approximate and can be subject to large errors, The number of particles sampled for each event is small, but the analyses carried out on each of these is detailed and within the errors shown, Standard deviations quoted in Table 1 are deviations of the mean value, not deviations of a single measurement from the mean.’ The radioactivity assay was done with counting instruments thoroughly calibrated and tested, Both standard isotopes and fission-product activities from the actual events were used to evaluate the instruments. The tabulation of activity per square foot (Table 2) reveals sampling biases which call for caution in the use of these data, Identical adjacent collectors, YAG 39=C-20 and YAG 39-C-24, sampled the same fallout event, yet the total activities recorded for Flathead for these stations differ by a factor of 5. Samplers of somewhat different design considerably o13. mon UNCLASSIFIED