-7= made to estimate the amount of radioactivity in storage and the rate of transfer from this reservoir by subtracting the total observed fallout from the total amount released in tests. Fig. 10 shows an example (Ref. 6). This method led to estimates varying from 1 to 3 megacuries of sr90 in the stratos- phere in the period late 1957 to early 1958 (Ref. 7, 8, 9). Uncertainties in the estimates of total yields of tests by all nations and in the estimates of integrated fallout over the earth prevent any greater precision in estimates by the difference method. A large part of the uncertainty is in the estimate of the fraction of "local fallout". This is the fraction contained in the region of maximum deposition within a few hundred miles from the test site. In Pacific tests, for example, it is virtually impossible to obtain a sufficiently complete map of the local fallout to permit accurate integration. In order to obtain direct measurements of the concentrations of strontiun90 in the stratosphere, a series of balloon sampling flights was made by the New York Operations Office of the AEC in 1953 and 1954 at about 80,000 to 100,000 ft. over Holloman Air Force Base, New Mexico with a simple electrostatic precipitator (Ref. 10). for flight. Fig. 11 shows a pair of sampling units ready This gave qualitative indications of the presence of Operation Castle debris in the summer of 1954 (Fig. 12). Quantitative radiochemical analysis was impractical, however, because the samples collected were too small and the efficiency of collection was unknown, In the fall of 1956, after extensive operational tests, a program of routine stratospheric balloon sampling flights devetopedbyGéeneratmMtits, Ire. was rockpeddeyBn the poeppretton of the Air Force using a filter ele \2 DOE ARCHIVES paper system, This program has had the nickname "Project Ash Can" because ~ of the shape of the sampling device (Fig. 13). Samples are collected at