showed that the slope of the line is not significantly different from zero. For a few low-yield airbursts, data have been obtained for single or grouped particles as small as 0.3 um. These submicrometer particles had been located on slide preparations of the debris by photoreversal of the autoradiographic image, transferred to an appropriate substrate, and Measured in an electron microscope. Activity measurements were made on groups of particles of approximately the same size. Size separation of debris by sequential centrifugation extended the size range downward even further, but also magnified the uncertainties of the measurements. Examples of gross activity behavior are shown in Figures 2 and 3. In Figure 2, the specific activity is essentially constant above about 2 um, then increases towards smaller particles sizes. The data from the size fractions appear to fit the single-particle data rather well, but our experience has shown that such data may have uncertainties as large as an order of magnitude or even more. This is evident, for example, from Figure 3, where the size-fraction data are a factor of about 6 lower than the single-particle data. In general, plots like these indicate that below a few micrometers, the specific activity curve begins to rise, and that this rise may become very dramatic below a few tenths of a micrometer. A word of caution is in order. The curves should not be blindly extrapolated below 0.04 or 0.05 um. Doing so would result in more radioactivity than was actually generated by the shot. In reality, there is a minimum particle size, probably of the order of 0.01 um, but the presence of a maximum in the specific activity-size curve also cannot be discounted. Evidence not adduced here suggests that the transition between constancy and increase of specific activity occurs at smaller particle sizes as the explosion yield increases. Individual radionuclides for which data have been obtained behave similar to the gross activities, as expected. However, the behavior above about 3 um is dictated by the relative volatilities and, therefore, the specific abundance of a particular radionuclide is not necessarily constant in this size range. Uranium and plutonium data are few and far between, but these elements may also be expected to behave in accordance with our general observations. Balloon and Low-Altitude Airbursts Data for this type of shot are rather scarce. We have analyzed data from King (Ivy series, high-yield device detonated at a scaled height of burst under 200 feet over coral, soil, and seawater) and from Lassen, Priscilla, and Wilson (Plumbbob series, devices with yields under 100 kt suspended from tethered balloons over alluvium with scaled heights of burst above 200 feet). 569