Document RjVLR3oO358ggN384K5qzd6bv

730 LEON F. CURTISS air, such methods are not generally reliable. Therefore,.the measurements milt be made in specially equipped laboratories by experienced technicians along lines now to be described. This requires that the samples of expired air be gto the laboratories in suitable flasks. The half-life of the- radon is sufficient | permit its shipment by air to any point in the United States for measuremenp'' there is no delay in handling. The activity decreases 18 per cent per 24-hour ;d and even two days after sampling a significant fraction of the original actiy remains. Measurement in a central laboratory has the additional advantage t all observations are more likely to be directly comparable than those mtf locally or in several laboratories under slightly different conditions. Howgy. radon standards, in the form of dilute solutions containing known amounts! radium, are available from the National Bureau of Standards so there is no j excuse for marked variations in measurements made at several laboratorle| apparatus of comparable sensitivity and reliability is used. ;i Laboratory measurements of samples of air containing radon are based! a comparison between the radon in the sample and a standard quantity of ra of similar magnitude. The standard quantity of radon is obtained from a stan solution of radium carefully prepared for this purpose. From the known ray accumulation of radon in such a solution, various quantities of radon cat obtained up to the equilibrium amount, by allowing the radon to accumulate :o., various periods of time. The growth of radon is given by the relation E( 1 -- e~-18!), where R is the amount of radon at any time t, measured in; pi since its removal from the radium source; e is the base of natural logarithms^! E is the equilibrium amount, equal in curies to the number of grams of rip| in the standard solution. About 30 days are required to reach practicallfe equilibrium amount since after this time e~-18* is for practical purposes equl zero. One half the equilibrium amount is obtained in 3.83 days so that vu fractions of'this amount may readily be obtained down to those so small; the accumulation time is difficult to measure accurately. The comparisons may be made by several methods, but in each the ionJf produced in an ionization chamber, by the alpha particles, is the effect custbMl, used. The number of alpha particles produced per hour, or other convip! unit of time, is proportional to the quantity of. radon in the chamber;, number may be determined by direct counting of the current pulses vinil^ chamber produced by the alpha particles, or by measuring, electrometer the current produced by the alpha particles traversing. thf$|p. the chamber at any time, which is indirectly a method of determining the,.flu of alpha particles emitted per unit of time. ---------En-additton-to-being-dependent-on--the-properties-ef--the-eqijijpnenl the accuracy of all such measurements is dependent on statistical laws,gpy| the evaluation of random events, since alpha particles are emitted by any,* i active source at a purely random rate. When this rate is low, as for the, quantities of radon, these fluctuations become obvious. This means that 4 RADIANT ENERGY 731 ^number of particles.must contribute to the measurement to reduce the standdeviation to the required percentage of the total number observed. .Since j^S-case the standard deviation is equal to \/N, where N is the total number Iffarticles recorded, an observation including 100 particles has a standard ration, on a statistical basis alone, of 10 or 10%. To achieve a. statistical SShVcurac-y of 1% it iB necessary to record 10,000 particles. Whether this con' ':jiis properly fulfilled is most conveniently determined in a method ;:iri which 'js.^e'ija'lpha particles are counted directly. The counting method offers other adfjgjffijptes that make it more convenient t'o use. If, to secure the desired statistical ^'^gliaicy, the time of observation must be extended over a significant fraction of Sfalfrlife of radon, correction must be made for the decay of the radon in the lift Double Ionization Chamber rlilH Calibrating Switch -4 3- String Electrometer 1 J-c> -B- To Pump Stondord Radium Solution Unknown Ensure 4. Schematic diagram of the essential elements of equipment for measuring samples dill by the use of a string electrometer to measure the ionization current produced by the whys from radon in an ionization chamber. S^' jfirr;''.nPr,,Furthermore, because, some, of the. immediate,decay products-iof,radon Sglyu.:twajlpffai^laiEiffies^and^FonwiButeTtorTKe^measurements',' oBSe'rviwidns "can be" ^KiguiWonly after-.these have come into transient equilibrium with the radon after V introduction into the ion chamber. This period is of the orderof three hours. KBflMabhematic diagram showing the: method18-'1?-of using afi'feredltbbneter .for 'Jl'V^tih'e collecting voltages of the two chambers opposed, is used, as shown, to ^Slize the effects of external sources of radiation. The radon is iiitroduCed gpiw. D. Urry and C. 8. Piggot, Am. J. Sci., 239, 633 (1941). ' "S D. Evans, Rev. Sci. Instr., 6, 99 (1935). .,