variable might also be included. If these distances and direction variables improved significantly the ability to predict plutonium concentrations in soil, then these variables might be incorporated into a double sampling plan to obtain yet more accurate and precise estimates of plutonium inventory. As indicated above, this has yet to be examined for its applicability. Gilbert et aZ. (1975) discuss the information available to date on plutonium amounts below the surface 5 cm of soil and within blow-sand mounds on the safety-shot sites. A closer look at the profile samples is needed and a comprehensive blow-sand mound sampling program is being planned. Considerable thought by several NAEG scientists has been given to how one might sample to estimate both the distribution and total amount of plutonium within mounds. Recommendations have been made but the problem is complex, and progress may be largely a trial and error approach. Concerning profile samples, an examination of present data strongly suggests that more detailed field notes should be taken at the time of sampling regarding the nature of the soil at the sample location (evidence of physical disturbance, erosion by wind or water, animal signs and burrows, etc.), both on the surface and at each depth in the profile. Information on the soil type (according, perhaps, to the classifications given by Leavitt (1974)) at each depth should also be recorded since soil type is a variable that may have an effect on the relative amounts of Pu in surface soil. This informa- tion should be recorded at each soil sample location. Consideration might also be given toward devising a field experiment to estimate the amount of cross contamination between different levels of the profile due to the sampling method itself. It is necessary to know how much of the plutonium being seen at depths is due to environmental weathering forces and how much to the sampling method. Such an experiment may be difficult to devise, but could be useful in the interpretation of these profile data. 106