8 LOCKHART AND PATTERSON 19.000 0.735 are artifacts due to the low accuracy in de- terming p (primarily a function of the total count during the 61-71 minute decay period) and that secular equilibrium did exist, the mitial count during the 1-11 minute decay period can be emRADON CONCENTRATION (pCi/m3)} ployed to calculate the radon concentration. Such calculations have been made for the extended series of collections made on January 5 and 1900 January 7, 1965; the results of both types of 100 measurement are compared in Figs. 9 and 10. The horizontal axis of the diamond-shaped pattern surrounding the points mdicates the collection time, while the vertical axis indicates the standard error (a7) in the determination of the radon concentration based on the uncertainty in he the measured activity ratio. The statistical errors associated with each of the measurements in the series which assumes equilibrium conditions to exist are not shown but are considerably smaller since the determination is based on a single measurement made during the period of highest 10 a | 5 | _ { 10 i) 20 STANDARD ERROR (PERCENT) activity. From Figs. 9 and 10 it may be concluded that no advantage ensues from use of the more compli- 4 25 Fig. 8 ~— Estrmations of the statistical error inherent in cated the determination of radon concentrations using filter samples counted through 10-mil aluminum absorbers assuming a 20-min sampling period at the rate of determinations of radon activity made by numer- 0.5 m/min, a background of 30 counts/min, and a 5% ous investigators who have mvariably assumed secular equilibrium to exist in the free atmosphere. contribution of ThB+C and fission products. having the rather large, of the measurements, random errors as- sociated with low counting rates. For values of p tn excess of 0.735 to occur in the free atmosphere, as is suggested by a few some mechanism procedure, at least under the sampling conditions existing here. Indeed, it permits one to place more reliance on the results of the many must exist there for removing the gaseous radon from its solid daughters or for fractionating the daughter products relative to one another. Wilkening (3) has demonstrated one such fractionation process in which RaA is preferentially removed by the strong electrostatic charge developed in the atmosphere dur.ng thunderstorms. This loss of RaA would interrupt the radon decay chain with the result that the RaC/RaB atom ratio (p) would Increase above the equilibrium value. There was no indication that such a mechanism was in operation during any of our sample collections; however, no electric field measurements were made. If the assumption ts made that the observed departures of p from the equilibrium value of It would appear that, in general, this approach ts valid. On the other hand, during the developing stages of a strong temperature inversion in the atmosphere, or immediately following ventilation of a mine, tunnel, or similar closed space, de- partures from equilibrium would be expected and the simple approach could sertously under- estimate the radon concentration. The validity of this approach for determining radon concentrations from measurements of Its filterable decay products was tusted by the chamber studies in which a measured quantity of radon was released into closed spaces of known volume as described earlier. The data accumulated in two experiments are summarized in Tables 5 and 6. In both instances the expected progressive in- creases in p with time were found. The experi- mental values of p obtained through use of the described procedure are compared in Figs. 1! and 12 with calculated results which duplicate as nearly as possible the actual conditions