SIZE DISTRIBUTIONS OF RADIOACTIVE AEROSOLS 161 for total beta measurement on miniature thin-wall Geiger counters whose backgrounds were about 0.2 count/min with 20-cm-steel and anticoincidence shielding.® Filters were measured for total beta activity on a larger Geiger counter of identical design with identical shielding with a background of about 6 counts/min. All total beta counting rates were normalized by multiplying by factors of 3.44 (miniatur counter and 3.09 (large counter) to correct for geometry (G). Beta. absorption differences between the two counters were considered to be negligibly small. For total gamma measurements (0.1 to 1.0 Mev), card-mounted stages 1 and 2 (2.5 cm in diameter by 1.3 cm in 8-cm Nal(Tl) crystal located within distance between the crystal surface samples and pelletized filters height) were placed on an 8- by 10 cm of lead shielding. The and the card-mounted sam- ples was about 0.6 cm. Filter counting rates were normalized to those of stages 1 and 2 samples by anincrease of 40% to correct for geometry differences. Samples were then dissolved from the tape and the filter by digestion with aqua regia in the presence of appropriate carriers. Individual radionuclides were separated and purified by standard radiochemical procedures. All beta-activity measurements for the de- termination of Sr, Zr, MOR, and '4Ce were made on miniature Geiger counters by using the Libby thick-solid-sample technique,’ and radionuclides were identified by either beta-particle absorption in aluminum or by decay. Corrections for counter background, geometry, self-absorption, and external absorption were made in converting counts per minute into disintegrations per minute. For “Mn measurements the 0.84-Mev gamma ray was counted, and all gamma-counting data were recorded on a multichannel analyzer. Individual radionuclide concentrations in units of disintegrations per minute per cubic meter [S.T.P. (0°C and 760 mm Hg)] were obtained by using volume data provided by GMI. RESULTS AND GENERALDISCUSSION Sampling data for the impactor—filter collections for the four balloon flights are given in Table 1. Although there are slight variations, the sampling-altitude intervals will be considered to be 9 to 15, 15 to 21, 21 to 27, and 27 to 30 km, and the particle cutoff sizes will be considered to be 0.15 and 0.02 in radius for stages 1 and 2, respec- tively. Table 2 lists our sampling dates relative to announced atmospheric nuclear tests whose fission yields were at least about 1 Mt. Time intervals between the last reported test of at least this size and our sampling dates are 132, 145, and 236 days for collections of March 1962, May 1963, and August 1963, respectively. In each of these cases,