DISTRIBUTION OF AIRBORNE RADIOACTIVITY 129 The activity levels existing in the air at the time the various collections for radiochemical analysis were made are shown in Fig. 5. Both the monthly average air concentrations of fission products re- ported by the Radiation Surveillance Network, U.S. Public Health Service, for Washington, D. C., and the results of various short-term collections made at NRL are depicted. The wide scatter in the values of the short-term collections demonstrates the extreme dependence of the ground-level activity concentrations on meteorological processes bringing upper tropospheric air down to ground level and on those cleansing the air of particulate matter. As in previous years a strong spring maximum in the concentrations of airborne fission products resulting from the winter—spring subsidence of stratospheric air was observed during 1963. The results of the radiochemical analyses of the seven collections made during 1963 are presented in Table 4. The wide variation in the distribution of radioactivity among the filters of packs collected at different times during all seasons of the year demonstrates that local meteorological conditions are of more significance than purely sea- sonal effects as a causative mechanism. Indeed, the greatest variation observed among the samples occurred during consecutive collection periods in August 1963. As indicated in Fig. 3, the particle-size ranges represented by the activity on the three filters are roughly as follows: filter 1 (IPC 1478) contains the bulk of the activity associated with particles around 1.0 u in diameter and some contributionfrom smaller particles; filter 2 contains particles in the range 0.3 to 1.0 1}; and filter 3 contains the remainder, primarily in the size range below 0.3... Over a wide range of particle sizes, approximately 75% of the activity should appear on the middle filter of the pack; consequently, nuclide ratios measured here should reflect the average of the collection rather than specific fractionation effects. The data in Table 4 substantiate this conclusion. The most likely evidence for nuclide fractionation would show up in nuclide ratio differences between particles retained by the initial and the final filters where the largest and the smallest particles, respectively, would be concentrated. The ratios of activity of each nu- clide relative to Sr as areference onthe top filter to the sameactivity ratio on the final filter, (K/*"Sr)jpc/(K/*Sr) toe g, are listed in Table 5. It should be noted that any inadequacies in counter standardization are eliminated by this method of presentation; therefore the results should be as reliable as counter stability and counting statistics allow. For the radioactivity determinations the standard deviation, 0, based on counting, has been kept below + 1%, except in.the case of ®*Sr, whereits lower activity and the necessity of obtaining it by difference has led to larger errors.