of 30 pairs of liquid and solid fractions, 20 of which are collected over a consecutive sequence of 2-minute intervals (designated 2-1, 2-2, etc., in Table 3.37) and 10 of which are collected over a Simultaneous consecutive sequence of 10-minute intervals (designated 10-1, 10-2, etc., in Table 3.37). For Umbrella, both sampling sequences were started by a radiation trigger at zero time; however, only a partial collection was obtained. Because of misalinement of the indexing switches by shock, the last two sampling heads in the 10-minute Sequence did not open and the last 10 collections of the 2-minute sequence were probably obtained without any vacuum being applied to the sampling head. Thus, although the analytical results for all collections are tabulated, only Samples 10-1 through 10-8 and 2-1 through 2-10 are considered truly representative. ALL AFI samples were sealed upon recovery and returned to NRDL for analysis. When recovered in the field, the first samples in both the 2-minute and the 10-minute series were heavily loaded with visible residue, which upon cursory inspection resembled pulverized coral. Upon arrival at NRDL, each sampling head containing its water reservoir (Figure 2.9) was first counted on the bottom shelf of a doghouse counter—a l-inch-diameter, 1-inch-thick, thallium activated Nal crystal canned in spun aluminum and installed inside a large lead shield into which samples as large as 18 by 22 inches can be placed (Reference 33). The shelf-tocrystal distance is 3 feet. The water reservoirs were then removed and the volume of the contents measured. ff the reservoirs were dry, 25 mil of distilled water was added. A 4-mi aliquot of the liquid phase was then counted in a well counter—a 1¥,-inch-diameter, 2-inch-thick, thallium activated Nal crystal containing a central well % inch in diameter and 114 inches deep sealed ina Vyg-inch spun aluminum can (Reference 110). About 150 ml of chloroform was used to dissolve the DMTfilters and to rinse out the sampling head. The solution and rinse were then filtered through an HA Millipore filter (a cellulose nitrate membrane having a controlled pore size of 0.45 micron). The residue was distributed as uniformly as possible over a circle 34 mm in diameter and was washed with additional chloroform to assure complete removal of DMT. The residue was dried overnight in a desiccator, weighed, then mounted on a plastic cap and counted on Shelf 5 of End-Window Gamma Counter 3 (an instrument similar to that used inthe EPG). After removal of the DMTfilter, the empty sampling head was again counted in the doghouse counter. Large amounts of residual activity appeared to be adsorbed on the bottom screen and aluminum walls and was extremely difficult to remove. Counts of the discarded chloroform filtrates containing the dissolved DMTfilters indicated that less than 1 percent of the activity was lost in the transfer to HA Millipore filters. A summary of AFI results is presented in Table 3.37. All counts have been reduced toa fifth-shelf end- window gamma counter response at 6 days after zero time. Conversion factors were obtained by comparing the doghouse counts of two AOC samples obtained aboard the DD-592 during Umbrella with the well counts and end-window counts of aliquots from the same AOC samples. The factor converting a doghouse count to a fifth-shelf end-window count is approximately 100 (the exact value varies between 99 and 103 for the period of 5 to 45 days). A similar factor converting the well counts at 6 days to end-window counts at 6 days is 0.27. Since the time of count can cause a significant variation in this factor, all other measurements have been brought to the time of this measurement using the observed decay curves. The decays for both the solid and liquid fractions of Samples 10-3, 10-8, and 2-11 are presented in Figures 3.225 through 3.227. The liquid fractions of the AFI samples obtained during Umbrella are first considered. Samples 10-1 and 2-1 have about equivalent amounts of water associated with them; since Samples 2-2 through 2-5 are dry, all this liquid must have been deposited in the first 2 minutes after zero time. The inside diameter of the AFI sampling heads is 3.55 inches; therefore, if uniform deposition is assumed, the collection corresponds to a rainfall of about 7 in/hr. Photographic evidence from Wahoo gives a maximum crosswind throwout radius of about 1,800 feet (Section 3.3.1); trajectories for downwind positions for both Wahoo and Umbrella are not expected to be much greater. Thus, massive deposition of water from such sources seems unlikely. All evi- dence indicates that base surge was in the neighborhood of the DD-592 at about 1 minute; however, the ship’s washdown must also be considered as another possible source of the liquid 331