straight for an appteciable time the log-log decay slope was determined graphically. The decay slopes for all of the samples are summarized in Table 3.7. They are tabulated for three time ranges: early times are up to 4 days after detonstion; medium times are 4 to 9 days; and late times are after 9 days. In the 9 to 30-day period the solid fraction decay was generally more rapid than the original sample. The ultrafiltrate decayed more slowly than the original. The colloidal fraction usually decayed more slowly than the ultrafiltrate, The decay curves of the various fractions diverged more for islemd shot samples than those for the barge shot samples, The solid fraction from island shot samples decayed et about the same rate as the originel slurry, while with the barge shot samples the ultrafiltrete decayed like the original slurry. These results are logical in view of the gross distribution of the gamma activity between the liquid and solid phases for the two types of shot. 3.2.2.3 Gamma Energy Distribution of Physical State Fractions Lead absorption curves were taken on samples for Shots 1 and 2. The curve for the 2-Ad4 sample on Shot 2 was taken at two times. Some of the curves are shown in Figs. 3.5 through 3.7. All fractions vere nor- malized to a count of .1000 at zero thickness of lead absorber for better comparison, The absorption curve of eech fraction was analyzed into three component energies and the percentage of each component was determined by weighting the "zero-alsorber" count rate of each component energy by the relative photon efficiency as taken from Fig. 2.1. The results are tabulated in Table 3.8. It may *e noted that the average gamma energy of the "colloidal" fraction was consistently higher for both surface island and surface water shot samples, whereas the solid and ionic fractions show large differences in relative amount of each component and average energy. This agair lends support to the argument that selective absorption occurred on the ultrafiltrate. The low energy components range from 145 to 180 kev, the medium from 320 to 485 kev, and the high from 1620 to 1620 kev, The fractions of the three "apparent" gamma energies from the solid fraction of Shot 1 sample (1-251.03) were similar to those for the original sample. In addition, the ultrafiltrate (ionic) fraction had a higher percentage of the highest energy gemmas then did the solid fraction, while the colloidal fraction had a still hipher percentage of high energy gammas, solid, The order of average energy was colloidal > ionic > The 1-251.02 sample fractions were somewhat different; both the decay and the lead absorvtion show very little frectionetion of gamma emitting isotoves between the solid and the ionic fractions. However, the comparison of average energies amceng the nhysical state fractions of any sample is not as reliable an indicator of fractionation as is the comperison of the percentage of the high energy component among the fractions, The latter depends uron the observed count at high absorber thicknesses while the former depends upon slopes extrapolated from 2 or 3 points, For the Shot 2 sample (2-A4) absorption curve of the solid 41