76 MORRIS LOEB 



successive transformations of radium, for instance, are ef- 

 fected by the expulsions of the a-particles and that these have 

 atomic mass: an atom of radium, therefore, contains a finite 

 number of them. As the transformations are atomic and 

 not molecular, Rutherford's application of the mathematics 

 of mass-action can mean but one thing: that the various rates 

 of transformation depend upon the chances of encounter 

 and relative positions of the particles within the atom. These 

 rates, however, as measured by the period of decay, vary from 

 thousands of years to a few seconds for the different educts, 

 and that irregularly in the order of transformation such 

 great differences could only be explained by an infinite number 

 of components, with large free paths, in other words, elec- 

 trons. It would then remain to be shown what caused a 

 certain great number of negative electrons to form an electro- 

 positive a-particle, and become expelled with great violence 

 from their surroundings. 



Naturally, the failure of an hypothesis to explain certain 

 facts does not invalidate the latter. Rutherford's brilliant 

 analysis of the curves of increasing and decreasing ionization 

 and the agreement observed with calculated results prove that 

 he is not dealing with mere fortuitous coincidences. Many of 

 his conclusions seem incontrovertible upon his premises; 

 but here again, the advocatus didboli must step in and ask 

 whether the premises are axiomatic: two of them appear to 

 me to be doubtful. (1) A curve of decay is based on electro- 

 scopic measurements upon the tacit assumption that the rays 

 sent out by that particular phase are always the same; but we 

 are told that both a- and /3-rays vary greatly in speed and 

 momentum, hence neither variety would show a uniform ion- 

 izing power; assuming that a substance did send out a-rays 

 for a long time, but that their velocities were gradually re- 

 duced, would not the ionization indicate a more rapid decay 



