time. Since the exact time of collection cannot be established with an accuracy greater than +/, minute, the characteristic decay curves for the two shots are combined for the correction of all IC collections (Figure 3.36). If these combined curves are accepted, the characteristic Shape of decay curves for collections madein the first minute resembles that which would be expected from samples deficient in short-lived or enriched in long-lived fission products. The decay shapes for samples collected during the next 4 or 5 minutes Show an apparently continuous Variation with time of collection, which could be duplicated by the addition of short- lived or depletion of long-lived fission products. After about 5 minutes, the base surge decays no longer exhibit any significant change with time of collection. In specific instances where a direct contribution from radioactive water is suspected (U 2.7 on Umbrella), the observed decay curve is intermediate between a base surge decay and a water decay. Contrarily, the water decay curves, while quite distinct from the base surge family, Show little variation with either time or location of collection. In general, the characteristic IC decays seem to depend primarily upon whether the deposition resulted from the base surge or from other Sources with secondary modifications dependent upon the time of collection. Although the location of collection necessarily affects the time of deposition, this single factor in itself has apparently little significance. The data available is not sufficient to demonstrate the suggested time dependence conclusively; however, it is sufficient to suggest that a more rigorous investigation of this phenomenon on future underwater detonations might be rewarding. Further evidence of the fractionation of Hardtack samples is found in the radiochemical analysis of a number of samples collected at various surface stations for both Wahoo and Umbrella and of a few cloud samples collected by LASL aircraft shortly after zero time (Reference 105). In summary, these analyses show evidence of extreme fractionation of certain radionuclides with respect to Mo’*. Zirconium, ruthenium, tellurium, and total rare earths showed little fractionation, but the nuclides with gaseous precursors exhibit considerable fluc- tuation. The base surge samples for Wahoo show Sr" enrichments greater than 20, whereas the Ba* enrichments of the samples are approximately a third of those observed for Sr*®?. Conversely, base surge samples for Umbrella are enriched in Sr*by factors ranging between 3 and 10, with Ba!" enrichments as great as twice those observed for Sr". Ocean water samples from both events were deficient in both Sr" and Ba‘ by factors as large as 2, whereas a crater sample from Umbrella was deficient in Sr® by a factor of 10. these results is given in Reference 105. Exact statementof all The change in relative Sr*® and Ba‘® fractionation reported for Wahoo and Umbrella may represent an example of fractionation of gaseous fission products at venting. The suggestion that the IC decays demonstrate a consistent change dependent upon time of deposition during the first minutes after detonation invites some preliminary speculation on possible fractionation mechanisms that might be operative during the early stages of base surge generation. The gaseous precursors of such radionuclides as Sr** and Ba'*® may not be dissolved in the water droplets comprising the plumes and base surge, whereas most nonvolatile fission products could be effectively scavenged either in the ocean prior to venting or subsequently by these same liquid droplets. So long as these precursors remain gaseous, they can exist independently, going into solution in the liquid droplets at rates that are slow in comparison to the rate of surge development. Upon decay to a nonvolatile daughter, however, the radionuclide would be strongly attracted to any available surface. Since in the column and early base surge a very large area of liquid surface exists, the rapid incorporation of these nonvolatile radio- nuclides can be presumed. Assuming that only a small percentage of the total fission products becomes airborne after an wkicrwater detonation but that a large proportion of the volatile products escape when the explosion hubble reaches the water surface and become mixed with the column and base surge, it seems reasonable to suppose that base surge droplets would become increasingly enriched with the decay products of the gaseous radionuclides. Assuming that various small percentages of the total fission products escape at bubble surfacing time, the calculated enrichment with 325