collected by the AFI. Washdown at first appears to be the only possible source for the iiguid fractions collected later than a few minutes after zero time. The data does not, however, support this supposition. Because of the position of the washdown spray nozzles relative to the platform, only a light spray could have reached the AFl—-a supposition that is supported by motion-picture photography of the Ships under washdown. If washdown were the source, then it should also account for the heavy deposition of radioactive material in the later AFT collections. To obtain active deposits of this magnitude, the water pumped through the washdown system would have to be about 700 times more radioactive than that actually observed in the vicinity of the DD-592 by the early recovery team, which boarded the vessel at about H+2 hours. Althcugh smalj areas of such highly radioactive water might exist, any contaminating event of this magnitude caused by washdown at these later times would certainly have registered as a pronounced rise in the GITR dose rate; no such large increase in dose rate was recorded. Although washdown may have made minor intermittent contributions, a more reasonable explanation of both the liquid and the solid fractions of Samples 10-6, 10-8, 10-9, and 10-10 is that the deposition occurred during the first minute after zero time when Samples 10-1 and 2-1 were collected. Base surge arriving at the DD-592 was traveling at a velocity between 40 and 50 knots (Section 3.3.4) and AFI Samples 10-10, 2-10, and 2-20 were close to the starboard edge of the platform, the side which faced surface zero. The impact of the base surgetore all the Project 2.2 animal cages along the starboard edge out of their mounting brackets and thus could easily have been sufficient to force considerable liquid into closed AFI sampling heads. Both rows of 2-minute sampling heads were shielded from the direct impact of the base surge by the starboard IC collector; however, the 10-minute sampling heads near the starboard edge must have received the full force of the impact. Thus, these later AFT collections are interpreted as additional evidence for a heavy liquid deposition associated with radioactive material occurring during the first few minutes. The specific activity of the liquid fraction is almost exactly the same for Samples 10-1 and 2-1 and is about the same order of magnitude for all liquid fractions except Sample 10-6 (Table 3.38). Not all AFI samples, however, show a consistent relationship between the total chloride content and the total activity of the liquid and solid fraction. A limited number of chloride analyses (Volhard) were run on Selected IC trays, and these show a similar iack of correspondence. Thus, although the individual radioactive droplets are undoubtedl. comprised of sea water, large amounts of nonactive Sea-water droplets appear to mask any possible relationship between total activity and chloride content. The specific activity of the solid material recovered from the DMT filter shows wide fluctu- ation, which may be attributed to the variable inclusion of nonradioactive solids and to the large amounts of activity lost to the walls of the sampling head. The specific activities are not, therefore, considered truly representative and have not been included in Table 3.39. Furthermore, when this solid material is leached with warm water, a considerable proportion of the total weight is found to be soluble. This soluble material, when crystallized, was petrographically identified as salt. Consequently, the approximate concentrations of sodium and chloride ions in the filtrate were determined by flame photometry and Volhard analysis respectively and are summarized in Table 3.39. The relative activity of the remaining solids and the filtrate could not be determined, since the separation was made too long after the time of detonation. The amount of solid material remaining after leaching could not be determined directly, because the preparation of the sample for particle size work did not permit accurate second weighings. This amount has, however, been estimated as a percentage of the original weight recovered from the filter by assuming that the sum of the sodium and chloride determinations approximate the total weight of material in solution. Since other ions known to be in sea water have not been taken into account, the percentages presented in Table 3.39 are probably high. Interpretation of the AFI data is made difficult bot’: by apparent sample bias due to the turbulence over the platform and by the high loss of activity to the walls of the sampling head. The material balance for an individual sampling head is good when wall losses are included; however, » yy 336