994 



EXPLORATION GEOPHYSICS 



elements (Figure 605). Therefore, if we compare the radioactivity of an 

 impermeable mineral such as pitchblende with the radioactivity of carno- 

 tite, the carnotite will appear low even though the uranium content of the 

 two minerals be the same. The ability of a mineral to lose radon is called 

 the emanating power and may vary widely from specimen to specimen. 

 This effect may be a serious source of error in radiological sample testing, 

 unless the samples are treated to fix the gas in place and then stored for a 

 few weeks to reestablish the radon equilibrium. 



Another important ratio is that 

 between JJ-^^ and radium, which is 

 usually constant in nature. There 

 are many instances, however, 

 where minerals are undergoing 

 leaching or chemical weathering 

 which results in differential move- 

 ment of the active elements. In 

 this way the equilibrium may be 

 disturbed faster than it can re- 

 establish itself and anomalous 

 assay results are obtained from 

 radioactivity measurements. Fig- 

 ure 606 shows the time required to 

 establish equilibrium between 

 f/238 ^j-,(j i^a^-6 'pj-,g question of 



the radium-uranium ratio in carnotites has been investigated by Lind and 

 co-workers,t who found the ratio constant in large samples even though indi- 

 vidual specimens showed various degrees of deviation from the average value. 



2 so 





100 200 300 400 500 6 00 



t, THOUSANDS OF YEARS — ► 



Fig. 606. — Uranium-radium equilibrium plotted 

 as a function of time. (After A. F. Kovarik, Nat. 

 Res. Council Bull. 80, 110, 1931.) 



Statistical Error 



The decay of a radioactive nucleus is not influenced by events that 

 preceded it nor by events which may follow. The radiations therefore are 

 emitted at random (or "perfectly irregular") intervals. 



In any random procession the number of events which occur in one 

 unit of time is not the same as the number of events in the following unit 

 of time, and the fractional difference between a single measurement and 

 an average of a large number of measurements is referred to as the "sta- 

 tistical error." This quantity will be different for each individual measure- 

 ment, and it cannot be given a definite numerical value. Instead, we must 

 think of it in terms of probability. The absolute probable* error of a 

 measurement is defined as 



/> = 0.6745 VN 



t S. C. Lind and C. F. Whittemore, "The Radium-Uranium Ratio in Carnotites," U. S. Bur. 

 Mines Tech. Paper 88, 29 pp. (1915). 



* The commonly used percentage probable error is the ratio of the absolute prob- 

 able error and the measured quantity, expressed in per cent (see Figure 607). 



