A problem arose in that most of the remaining sample aliquots from TTR
and Area 11 sites that had been shipped to the analytical laboratories
for analysis were of <100-mesh soil, i.e., these aliquots were of soil
that had passed through a 100-mesh screen after ball-milling.
But the
results reported in Gilbert et al. (1975) were for aliquots of only
ball-milled soil, i.e., for which no sieving was done.
The question
arose as to whether results on <100-mesh and unsieved soil are really
comparable, i.e., do these two soil fractions contain (on the average)
different levels of Pu concentration?
Gilbert and Eberhardt (1976a)
examined the available data and concluded that <100-mesh soil aliquots
in relatively high activity soils tended to have higher concentrations
than unsieved aliquots, but that the effect tended to be less pronounced
for low-level samples.
It was decided to circumvent the problem by
preparing new unsieved aliquots from stored ball-milled library samples
to replace the <100-mesh aliquots previously shipped to the analytical
laboratories.
The Pu and Am (Ge(Li)) results on these new unsieved
aliquots were reported to us in May of 1977.
However, our attention to

other NAEG studies (primarily the blow-sand mound studies) and reduced

levels of support since October, 1977, have prevented the statistical
analyses of these data.
This is most unfortunate since the major expense
of chemical analysis has been completed.
Additional information also exists from 188 soil and 173 vegetation
samples collected at new locations from all 10 safety-shot sites to ©

"fill in the gaps" left by the original random selection process.

We

requested that these new samples be collected so that improved estimates
of Pu spatial distribution (discussed below) and inventory could be
obtained.
The Pu and Am concentrations on unsieved 10-gram aliquots of
these samples were also reported in the summer of 1977.
These data have
been placed on computer cards in preparation for statistical treatment,
but the actual analyses have not begun due to budget restrictions.
These new location data and the data on unsieved aliquots discussed in
the previous paragraph have been entered into the NAEG data bank by
Reynolds Electrical & Engineering Co. (REECo) personnel.
In the spring of 1977, we discovered that some estimates of Pu inventory

for safety-shot sites reported in Gilbert ef az.

(1975) were in error

due to inaccurate planimetering of the Am activity maps (Figures 4-14,
Gilbert et al., 1975).
Subsequently, the size of all strata was recomputed by dividing each stratum into small squares and counting the number
of squares falling within the stratum.
These new stratum sizes were
then used to recompute Pu inventory estimates which are reported by
Gilbert

(1977b).

The corrected total estimated inventory for the

10 sites is about 146 curies as compared to a total of about 155 curies

originally reported by Gilbert et az.

(1975).

454

Select target paragraph3