150 ANNUAL REPOET SMITHSONIAN INSTITUTION, 19 3 2 



million years. If, however, we observe that it bursts at a time t 

 we can ascribe to the atom the retrospective character Kt^ meaning 

 that it had (all along) the property that it was going to burst at 

 time t. Now according to modern physics the character Kt is not 

 manifested in any way — is not even represented in our mathematical 

 description of the atom — until the time t when the burst occurs and 

 the character Kt having finished its job disappears. In these cir- 

 cumstances Kt is not a predetermining cause. Our retrospective 

 labels and characters add nothing to the plain observational fact 

 that the burst occurred without warning at the moment t; they are 

 merely devices for ringing a change on the tenses. 



The time of break up of a radioactive atom is an example of ex- 

 treme indeterminism ; but it must be understood that according 

 to current theory all future events are indeterminate in greater or 

 lesser degree, and differ only in the margin of uncertainty. When 

 the uncertainty is below our limits of measurement the event is 

 looked upon as practically determinate; determinacy in this sense 

 is relative to the refinement of our measurements. A being accus- 

 tomed to time on the cosmic scale, who was not particular to a few 

 hundred million years or so, might regard the time of break up of the 

 radioactive atom as practically determinate. There is one unified 

 system of secondary law throughout physics and a continuous grada- 

 tion from phenomena predictable with overwhelming probability to 

 phenomena which are altogether indeterminate. 



The statement that all phenomena haA^e some degree of indeter- 

 minacy will probably be criticized as too sweeping. I will consider 

 just one example. I have said that a Kzo atom is not radioactive. 

 Then (it will be said) we can at least state one predetermined fact 

 about its future; we can predict without any indeterminacy that it 

 Avill not break up as a K^r atom would do. The answer of modern 

 physics is that strictly speaking there is no such thing as a /f 39 atom, 

 but only an atom which has a high probability of being ^"39. Such 

 an atom should contain 39 protons within a small nucleus; but the 

 proton in modern physics has a very important peculiarity, viz, it 

 never is anywhere quite definitely though it may have a greater 

 probability of being in one place rather than another. Thus we can 

 never get beyond a high probability of 39 protons being collected 

 together. It is impossible to trap modern physics into predicting 

 anything with complete determinacy, because it deals with prob- 

 abilities from the outset. 



It has seemed necessary for clearness to give an example of an 

 event believed to be widely indeterminate; but you must not sup- 

 pose that I have brought forward the phenomenon of radioactivity 

 as evidence for indeterminism. There is a widespread idea that 



