d in surface soil, where relatively high contamination levels foun those for BM <100 mesh results tend to be considerably higher than most Second, the ratio of BM to <100 mesh is not, in fractions. Concerning Question n. cases, eonstant over all levels of contaminatio ns from 2, the error in estimating inventory using Pu concentratio o of concentra<100 mesh rather than BM aliquots depends on the rati rions in these two soil fractions which (as noted above) is not Probably a conservative constant over all concentration levels. the Pu data in estimate of the average effect can be obtained from These are surface samples from the second stratum of Area Table 6. 13, a low-level stratum as compared to strata closer to GZ at that 0.5 site. For these data, the average ratio (sieved/BM) is 2.6 + (X + S.E.) so that the estimated inventory would be elevated by a factor of 2.6. For strata nearer ground zero, one sees from Table 4 for soil samples in the top 0-2.5 or 2.5-5.0 cm in stratum 4 (the highest activity stratum for the TTR sites) that the ratio can be expected to be closer to 15 or 20. Hence, the estimated inventories for these high-activity strata would be similarly elevated. Turning to Question 3 as to whether Pu concentrations on sieved samples can somehow be corrected back to BM fraction or "field" conditions, the data in Table 8 is presented for consideration. These are profile samples from the TIR (a subset of those in Table 4) for which 239-240 Pu concentrations were obtained on BM, <100 mesh and >100 mesh aliquots. After the 10-gram BM fraction was removed from the total BM sample, as much as possible of the remaining ballmilled soil was passed through a 100 mesh screen, and the weights of the total <100 and >100 mesh fractions recorded. A 239-2405, concen- tration "corrected" to field conditions can be obtained by computing 8 = b(ata)t eGra)- tf (1) where the letters correspond to the labeled columns in Table 8; i.e., a= weight of <100 fraction, d = weight of >100 fraction, b = uncorrected Pu concentration in <100 fraction, e = uncorrected Pu concentration in the >100 fraction, c = Pu concentration of <100 mesh corrected to field conditions, and f = Pu concentration of >100 mesh corrected to field conditions. Hence, g is a "weighted" average, where a/(atd) ~ proportion by weight of <100 mesh soil, and d/(atd) = proportion 151