The results are converted from activity per gram of sample analyzed (less than 10-mesh material) to activity per unit area by multiplying by the mass of sample (less than 10 mesh) per area sampled. The upper 5-cm samples from Site IV contained hard clods (up to about 5 cm3) that could not be broken by autogenous milling or by hand. Based on observations of the actual presence of rocks and the amount of mate- rial that readily broke up and additional material that was broken up in a selected sample, 90 percent of the sample was assumed to be less than 10 mesh. RESULTS AND DISCUSSION Table 1 presents the plutonium-239 results for the top i~cm and top 5-cm samples for the four sites. sample and nCi/m2. The date are given both as fCi/g of The results are for the less than 10~mesh (2-mm) soil fraction. Most of the subsequent data treatment is based on the activity per unit area, thus the average, standard deviation, and coefficient of variation (C) are given for the nCi/m2 values. The C for the activity per gram values is somewhat smaller than that for the activity per unit area. Similar data are given in Table 2 for the special samples and depth profile (5- to 10-cm) samples. The sites were sampled and the samples prepared in order of increasing expected activity (Sites I, III, IV, and II, respectively) to minimize the hazards of cross-contamination. Appendix 1 includes a brief descrip- tion of the site characteristics. 0.1 to 1 pCi/g. The expected range of activity was With the exception of two values, the sample results reflected the expected values and, recognizing the difference in sampling depths, are similar to the isopleth values given in Figure 1 after Krey and Hardy, 1970). deposition down to 20 cm. (DOE, 1977, Krey and Hardy's values are for the total It took about the same amount of time for collecting samples by both techniques. The individual who collected the samples for the O- to 5-cm technique had previously used the technique. All of the 0- to 5-cm samples were collected by the same person. The individual who collected most of the O- to l-cm samples had not previously used the technique in the field. Most of the O- to l-cm samples were collected by the same individual, but about 20 percent of the samples were collected by a second person. The result for sample III-17 is grossly higher than any of the other values, and is especially higher than the other values for Site III (Tables 1 and 2). Recounting the sample confirmed the indicated value. A new aliquot of the sample is being analyzed. It is difficult to speculate how this degree of sample contamination would have occurred, 652