the alluvial soil. However, this relationship between particle density and soil density does not seem to hold for surface bursts. The particles from the balloon shots have an anomalously low density, presumably as a result of vesiculation. The densities of the particles from the underground shot appear to increase as they become smaller. This is consistent with an expected decrease in vesiculation with decreasing particle size, We conclude that for large particles (d > 100 um), the density is about 2.4 g/cem?, but that each shot generated particles with considerably lower and higher densities. THE PARTICLE SIZE DEPENDENCE OF THE CONCENTRATION OF TRANSURANICS Introduction The extent of the data has been discussed in a previous paper. It was found that until the early sixties, mostly gross activity measurements in fallout samples were made. Sometimes these samples were size-fractionated by means of sieving. At other times, fused particles were isolated and their radioactivity measured. Measurements of the concentrations of individual radionuclides were rarely made in size-separated samples or in individual particles. The principal characteristics of all these measurements were that they were made rather shortly after each test and that they covered the largest particles of our interest in the fallout field. Primarily as a result of the pioneering work of Russell (1965) and of Heft and Steele (1968), more detailed work on both fallout and cloud samples was performed. The investigations covered not only the debris from tests conducted during the sixties, but also samples from earlier tests in the Pacific and at the Nevada Test Site. Only cloud samples above about 200 um, and in some cases less, could not be considered. We did not find cases where an attempt was made to isolate individual particles for measurement of these radionuclide concentrations. In this particular report, we will discuss only events that occurred in the very early sixties and before. The concern for potential health hazards from radioactive particles at the Test Site requires us to look into the activity of individual particles, and the problem is: Can we infer the concentrations of transuranics, or at least the particle-size-dependence of these concentrations, from the type of data that are available, particularly from beta and gamma measurements? This problem is attached in the next subsection. 561