540 SCIENCE PROGRESS 



The average life of a radioactive element at any definite time 

 corresponds to the first of these illustrations. If we take a 

 definite quantity of radium salt, then the atoms of radium it 

 contains are of all ages. If they have just previously been 

 separated from their uranium ore some may have been born only 

 a few weeks previously, others have existed thousands of years, 

 just as the church congregation may include infants and old men 

 of ninety. Of the radium atoms a certain number will expire 

 in a definite time, just as in so many years a number of the 

 worshippers will have died. And we can calculate the average 

 life of the congregation of atoms with as much certainty as 

 that of people. 



The application of the second kind of average life to radio- 

 active substances was recently worked out in some detail by 

 Sir William Ramsay as a lecture illustration. Though it has 

 just been stated that the life of a radioactive atom is infinite, 

 yet infinity is only a relative term ; in this connection it only 

 means a longer time than we can measure. It is extremely 

 probable that just as the human life is measurable and definite, 

 so is, let us say, the life of an atom of radium. One may 

 go further. For there is much more regularity in the life of 

 this dead matter ; the life of human beings is variable, dependent 

 on "all the ills that flesh is heir to," microbes, the bearers of 

 disease, and accidents, the ills of fortune. We have no ground 

 to suppose that the life of a radioactive atom varies from any 

 similar cause. For example, the rate of decay — the death- 

 rate — of radium emanation is unaltered within the limits of 

 measurement between the temperatures of — 180 and+ 1600 C. 

 (90— 1870 in the absolute scale), the extremes at which, so far, 

 it has been experimentally tested. If this tremendous difference 

 of temperature does not hasten the decay of a single atom of 

 emanation we can at least postulate that its life is not easily 

 influenced by external causes. So far experimental work 

 has shown that we cannot hasten or retard the death of a 

 radioactive atom ; consequently we cannot alter its life. For 

 its birth is but the death of its parent atom, to which the 

 preceding statement likewise applies. And we are justified in 

 stating that the average life of a radioactive atom, as used in 

 the second sense above, is in all probability the actual life of 

 that atom. 



