a comparison of the 2-minute and 10-minute series is poor, being particularly so at later The AFI data does indicate that large amounts of water were associated with the activity 1 base surge during the first minute, after which time the base surge appears to be compos principally of small droplets or nearly dry aggregates of salt slurry. This later stage is gested by laboratory tests of the DMT filter, which indicate that the filter retains less tha percent of the salt contained in clean sea water or in a suspension of Umbrella crater mat in sea water. The possibility that the activity in the later AFT samples is simply a gener: background due to very fine droplets accompanying the initial base surge exists, but it 1s sidered remote. Micrescopic examination of the solid residue from the AFI samples indicated that the s distribution determination originally intended would be largely useless. The material ran between 10 and 80 microns in diameter and appeai ed under the petrographic microscope t largely CaCO, and crystalline salt with a few opaque particles of iron oxide and some silic material probably not of device origin. Only one classical iron sphere (Reference 18) wr: served; this single sphere was not, however, radioautographed. Spectrochemical analysi this material confirmed the petrographic findings and also indicated the presence of other ments usually reported in fallout material (see Table 3.40). As demonstrated by the theoretical considerations in Reference 90, the size distributio the liquid drops comprising the base surge is undergoing large and rapid changes. The du of the surge is so short that it is unlikely that equilibrium conditions are attained before d tion. Furthermore, the ambient conditions within the base surge are not precisely known fore, their effect on the ultimate particle size distribution cannot be precisely determined only instruments that could give some indication of droplet size distribution were the two mental collectors installed aboard the DD-592, which were loaded with trays containing s| reagent films and a few vertical panels containing similar reagent films (Reference 111). fortunately, the amounts of water accompanying the base surge were so large as to rende: reagent films unreadable. Limited size information was obtained from a numberof the vertically mounted cellulo acetate sheets installed aboard the DD-593 only. These samples were recovered and retu to NRDL for analysis. Size data was obtained from these sheets using the isopiestic tech: desrrihed in Reference 112. This procedure is based on the fact that hygroscopic sea- sa} clusters will absorb water, if the relative humidity exceeds 75 percent, until the resultinj solution reaches a dilution at which the vapor pressure of the drops and the humid atmosp are equal. Thus, the acetate sheets were placed in an atmosphere maintained at approxir 80-percent relative humidity until the impinged salt particles had absorbed sufficient wate form little hemispheres of solution in equilibrium with ambient conditions. By measuring volume of these equilibrium hemispheres, the weight of salt deposited on the sheet and he the equivalent sea-water droplet diameter could be determined. The results of this analy showed a few large droplets of about 0.5-mm diameter with the preponderance of the drop diameters being 20 to 40 microns. According to the mathematical analysis in Reference | this type of collection would be very inefficient for the collection of droplets less than 101 crons in diameter; therefore, the actual median size for the droplet population mayhave t considerably less than 20 microns in diameter. A radioautograph of these sheets showed most droplets to be weakly radioactive. X-ra diffraction of the solids included in these droplets revealed the presence of CaCO, as both calcite and aragonite for Wahoo but only as aragonite for Umbrella. evidence of quartz and calcium silicates on both shots. Diffraction also shov A few water-insoluble crystals (b 10 microns) that appeared more active than the rest of the material were also found on bo shots. No radioactive iron spheres were observed in these samples. The apparent abser the iron spheres typically associated with tower or megaton-range barge shots should be: A few large coral grains were included in the Umbrella droplets, and a number of weakly active filamentous structures were observed on both shots. These fibers were generally uble in dilute nitric-hydrochloric acid mixtures but readily soluble in hot nitric-perchlor 339