of plutonium in the soil (both horizontal and vertical) plus the uncertainties of aliquoting and analysis of samples contribute to the variability of results over the uncertainties of the actual sampled mass. Except for the very low C indicated for the 0- to 5-cm technique for Site I, the C's for the O- to l-cm technique are essentially the same or lower than those for the 0- to 5-cm technique. Site I was the first site sampled, and from a petrological viewpoint, an ideal site for sampling. Site ILI was a very rocky hillside and Site IV a dry, hard-packed, highly compacted (silt to clay valley fill) pasture. Thus, the sampling scoops for the O- to 5-cm technique were in a bad state of disrepair prior to sampling Site II (sampling sequence I, III, IV, and II). It is noted that the variability in the upper 5-cm technique may reflect the inappropriateness of the scoop technique, which was developed for sandy desert areas, for the rocky and hard-packed loam soils around Rocky Flats. Rockwell International uses a steel 10- by 10-cm template-type device that is designed to be driven into the ground to define the sample. A field study with the scoop technique in sandy soil indicated an uncer- tainty of less than 10 percent in the mass per unit area sampled (Bernhardt, 1976, p. 20). The study was based on eight depth profiles to a depth of 20 cm by two different sampling teams. The O- to l-cm technique generally showed uniform variability (similar C's) at the various sites. The greatest variability was at Site IV, the hard-packed site. The ground hardness at this site was such that even the l-cm-deep template had to be driven into the ground with a hammer. Therefore, the increased variability is not surprising. Figure 7 illustrates the amount of plutonium-239 deposition per sampling depth denoted in this study. The results from cesium-137 are also shown for comparative purposes (the curve is given only as an approximation). Krey and Hardy (1970, p. 22) sampled to a depth of 20 cm to assess the total plutonium inventory around Rocky Flats. They noted that as much as 60 percent of the total activity (activity per unit area) was below 5 cm. Their results from seven sites indicated that an average of 62 percent (range of 39 to 91) of the total activity fell into the top 5-cm interval. Based on three sites, 93 percent (88-98) of the activity was in the top 10-cm interval. These values indicate 67 percent of the activity in the top 10-cm interval is in the top 5-cm interval. This is somewhat lower than our average estimate of 86 percent, but the range of values overlaps. Results from EPA samples in October, 1976 (Bernhardt et alZ., 1977), from a hillside site about 600 m north of Site II, indicated about 44 percent of the activity from the top 5 cm was in the upper 1 cm, which is compar- able to the average of 35 percent (range of 23 to 62) indicated by this study. 662