by weight of >100 mesh soil. One may think of ba/(atd) and ed/(at+d) as the <100 and >100 mesh fraction Pu concentrations, respectively, that have been corrected back to field conditions by taking into account the proportion of <100 and >100 mesh soil in the total soil sample. [Note that at+td as reported to the authors are slightly less than the total BM fraction weight (98.0 + 0.62 percent). The assump- tion is made here that no bias is introduced into the corrections by this fact. The total dry weight (before ball-milling) is given in Table 8]. The BM fraction 739-240 Pu concentrations (Column h) are plotted against "total corrected" concentrations (Column g) in Fig. 18. The resulting estimated linear regression line may be compared with that in Fig. 17 where these same BM fraction data are plotted versus the uncorrected sieved Pu concentrations (Column b). That is, by correcting the <100 and >100 mesh Pu concentrations to field conditions, does one obtain concentrations equivalent or closer to those obtained on the BM fraction? If the BM and "total corrected" concentrations are indeed “equal,"' then the two nonparallel lines in Fig. 17 should become parallel and coincident in Fig. 18; i.e., one should find By ~ 0 and B, * 1 in Fig. 18. One sees that Bo does get closer to zero (-0.355 changes to -0.252) and B, is closer to 1 (0.744 changes to 0.817), but By = -0.252 and By = 0.817 are still significantly different from 0 and 1, respectively. Hence, the "corrected" concentrations obtained using Equation (1) above did not, at least in this instance, fully correct to BM fraction results, although some improvement was noted. This raises the question as to whether differences between laboratories might be partly responsible for the lack of agreement between BM and sieved results in Fig. 9. The "corrections" applied above would not remove any such laboratory bias. 152