originally present in the device as the fuel, and little is formed by neutron capture by uranium; and like a nuclide of volatility intermediate between !4%Ba and 147Na, if both induced and original plutonium are present. We will investigate this matter further in the following subsection using data obtained from size~separated samples. We also determine to what extent plutonium behavior can be inferred from gross-activity behavior. Relation of Pu Behavior to Gross Activity Behavior The principle of our investigation is the following: Data on uranium, plutonium, and fission product concentrations in size-separated samples exist. These data relate the radionuclide concentrations to the particle size. Because of the age and nature of the samples, the fission products measured are usually limited to 90sr and !47Pm (147Nq-daughter), although in some samples, other fission products can also be determined. The fission product 147pm is a refractory nuclide. Such refractory nuclides are by far the most significant contributors to the gross beta and gross gamma activities. If the particle-size-dependence of plutonium is approximately the same as the particle-size-dependence of promethium (or other refractory species), such particle-size-dependence can then be inferred from the particle-size~dependence of the gross activities. The contribution of refractory nuclides to the total fission-product activity can be derived from the work of Bolles and Ballou (1956). In Table 4, we have shown the contributions of various groups of fission products to total activity at various times after a test. No distinction is made between different fuels. At about the time that the gross activity measurements were made (certainly less than a year after each test), the refractory nuclides contributed between 60% and 85% to the total fission-product activity, exclusive of the rare gases and the halogens. In addition, some fraction of the alkaline earths and noble metals was not incorporated in the fallout because of their own volatility or that of their precursors during fallout formation. Of the alkaline earths (Sr, Ba), perhaps the major portion is missing. If it is assumed that only 50% of the alkaline earths and noble metals were present in the fallout samples, only about 75% of the total fission product activity was represented in the fallout at 25 days and about 90% after 1 year. Therefore, the refractory fission products constituted from 80% to 954 of the fission product activity in the fallout during the time that measurements were made. Since the data are most often reported as the activity at 25 days after the test, the lower percentage figure would apply more generally. In order to obtain a comparison of the distribution of plutonium with that of refractory nuclides over different particle sizes, it is suffi- cient if the relative concentrations in size-separated fractions of cloud and fallout samples are known, even though particles of all kinds are usually present in the samples. In our analysis, it is assumed that sa :