used in earlier work in this Jaboratory.1l/ Its medium was a cellophane dialysis membrane previously found to have a pore size of 12 to 40 A, One thickness of cellophane at a pressure of 400 psig (under nitrogen) gave a flow rate of 3.5 to 4.0 mli/hr. After the whole supernatant aliquot, had passed the filter, a garma counting aliquot of the effluent was taken. Upon disassembly of the ultrefilter, the membrane was counted to determine the gamma activity in the collcidal fraction. The solid fraction separeted by centrifugation was transferred quantitatively to a weighed fritted glass filter using anhydrous methanol and after drying the filter was weighed again. Then the solid on the filter was dissoived with 6N-HCl and washed through. After several washings, the combined filtrates were transferred to a 100-n1l volumetric flask and made un to volume. An eliquot of this dissolved solid frace tion was taken from this solution for gamma counting. Thus, for each sample five gemma counts were taken, (1) original slurry, (2) supernatant, (3) ultrafiltrate, (4) colloidal fraction, and (5) solid fraction, This procedure allowed the calculation of an activity balance for the two separation steps. The last. three counts gave the breakdown of the gamma activity into fonic, colloidal, and solid fractions, These samples were also used to follow the gamma decay of the fractions for all the shots. For the fractions of the samples from Shots 1 and 2 lead absorption date were taken; for the fractions of the samples from Shots 2,3, and 4 ganma analyzer data were taken. Finally, portions of the fracticns of samples from Shots 2,3, and 4 were returned to USNRDL for quantitative analysis of their mejor and minor constituents, 3.2.2 Resulis The results of these studies are siven in the following sections. 302e2.1 Gamma Activily Distribution Among Physical State Fractions In general, gooc activity balances were obtained for the sera- ration steps. The sum of the total gcmma counts of the liquid (super- natant) and the solic fractions was 94 to 103 per cent of the total gemma count for all samples, except one, as determined by the assay of the original sample. Similarly, recoveries in the ultrefiltration step (sun of collcidal and ultrafiltered fractions) ran about &6 to 96 per cent of the total liquid activity. Totals were normalized to 100 per cent by taking account of the known sources of loss. In the solid~liquid separation the main source of loss was in the transfer of the solid to the frit and in the residue left on the frit after the acid wash. In the ultrafiltration separation the main source of loss was in adsorption on the metal surfaces of the ultrafilter below the membrane. Separate experimente showed that the extent ofthese losses was sufficient to account for an occasional low recovery. Tables 3.5 and 3.6 summarize the gamma activity fractionation results together with pH values and percentage of solids by weight. Table 3.5 gives the results for the individual samples while Table 3,6 gives ranves of values for all semples analyzed in each shot, and grouns these results by type of shot. Sone pertinent observations based on Table 3.€ are: 34