t Conversion of gamma intensity contour data to fraction of device requires knowledge of the relation of dose rate to fissions per unit aréa ofthe fallout field at 1 hour and of the gross radioactive decay rate. The decay rate varies with the device composition, environment, and fractionation in a way that is not well understood. Some uncertainty will always be present in local fallout determinations by this method when fractionation exists to an unknown degree, even though all the other quantities are known accurately. Another procedure for the determination of fallout partition was originated by the University of California Radiation Laboratory (UCRL) based on the supposition that certain of the rare-gas fission products remain throughout their lifetimes as free atoms unattached to surfaces (Reference 29). If this is true, they will not be removed from the cloud by the falling particles and may be considered as representative of the number of fissions remaining aloft for long periods. In the application of this method, coincident samples of gas and particles are taken by an isokinetic collector during the first few hours of existence of the clouds. The nuclear aerosol is sucked through a filter to remove the suspended material and the particle-free gas is then pumped into a storage bottle. The number of fissions in the two sampies is determined by analyzing the gas for 2.8-hour Kr*® and the solid for a representative nuclide such as Mo’. The ratio of sample fissions calculated from a bound nuclide to those from an unattached rare-gas nuclide will give the fraction of the reference substance that is in the sampled portion of the cloud at the time of sampling. Ata very early time, tf no separation of gas and particles occurs, this ratio should be 1. Later it would be expected to decrease as the falling particles remove the bound fission products. Hence, if the early ratio is 1, the fraction of the material in worldwide fallout may be determined if the time is known at which particles having a falling velocity of 3 in/sec leave the sampling region, or if the ratio approaches a constant with time. 1.2.5 Prior Estimates of Local Fallout. Determinations of local fallout have been made at virtually all the nuclear tests conducted by the United States. Estimates of the fraction of the radioactivity deposited locally have been made for Operations Jangle (References 17, 24, 28, 30, and 31), Tumbler-Snapper (References 17 and 30), Upshot-Knothole (References 17 and 30), Castle (References 32 through 36), Wigwam (Reference 37), Teapot (Reference 38), and Redwing (References 24 and 39). A summary of fraction of radioactivity deposited, computed from gamma contours and/or area sampling, covered a range from 0.2 to 0.6 (References 28 and 29). Reexamination of the preliminary Redwing data (Reference 40) gave higher figures in the range 0.65 to 0.70 for barge (water-surface) shots and up to 0.85 for land-surface shots. Results by the UCRL cloud-sampling method are also available from Operation Redwing (Reference 29) for the ground shots, Lacrosse, Mohawk, Zuni, and Tewa (part land, part water); for the water-surface shots, Huron and Navajo; and the high-altitude airburst, Shot Cherokee. In the first three events the ratio of solid-to-gas fissions was as low as 0.04. Values for Tewa were not much less than 1, but this was probably due to the low sampling altitudes relative to cloud height. cases. of 1. The ratios for the barge shots were greater than 0.6 in all For Shot Cherokee the only sample taken from the main body of the cloud gave a ratio From the assumption that the ratio at early times in all cases is 1, interpretation of these figures in'termsof fallout distribution indicates that 90 to 95 percent of the activity came down locally for the land shots, 15 to 50 percent for the water shots, and essentially none for the high-altitude airburst. On 5 to 7 March1957, a symposium was held at The RAND Corporation to summarize and evaluate work done on fallout partition up to that time (Reference 29). The conferees concluded that the best generalization that could be reached on the basis of the data presented was an equal distribution of radioactivity between worldwide and local fallout for both land and water detonations in the megaton range. 1.2.6 Worldwide Fallout. Worldwide fallout has been of great concern to persons responsible for the conduct of nuclear tests because of the possible consequences attendant upon the global dispersal of radioactive substances (References 41 and 42). The dangers from external irradiation are generally believed to be of a minor nature because of the low levels of activity 16