54 



Professor J. J. Thomson 



[March 10, 



would produce transmutation of the elements. Whether the atoms of 

 our elements will tend to break up into the atoms of other elements 

 will depend upon the relative stability of the atoms, and the stability 

 of an atom will depend mainly upon its potential energy ; if this is large, 

 the atom will be liable to break up or change. I have calculated for 

 atoms containing from 1 to 8 corpuscles the potential energy of the 

 atom per corpuscle : i.e. the potential energy of the atom divided by 

 the number of corpuscles in the atom, making the assumption that 

 positive electricity behaves like an incompressible fluid, i.e. that its 

 density is invariable. The result is represented graphically in Fig. 3 ; 

 the vertical ordinates represent the potential energy per corpuscle, the 

 horizontal abscissse the number of corpuscles in the atom. You will 

 notice that the curve is a wavy line with peaks and valleys ; the atoms 

 corresponding to the peaks would have greater potential energy than 

 their neighbours, and would therefore tend to be unstable, while those 

 in the valleys, having relatively little potential energy, would be stable. 



Fig. 3. 



The case is in many respects very analogous to the case of a number 

 of stones scattered over a hilly country whose section is represented 

 by Fig. 3 ; the stones, if subject to disturbances, would run from 

 the hills into the valleys, and though the stones might be uniformly 

 distributed to begin with, yet in course of time they would accumulate 

 in the valleys. So also in the chemical problem, though the number 

 of atoms of the different elements might initially not be very unequal, 

 yet, in course of time, those in the valleys would increase, and those 

 on the peaks diminish, so that some elements would increase, while 

 others would tend to become extinct. The smallest potential energy 

 is that of an atom consisting of a single corpuscle ; this is the goal which 

 all the atoms would ultimately reach, if subject to disturbances 

 sufficiently intense to lift them over the intervening peaks. Thus, on 

 this view, the general trend of the universe would be towards simpli- 

 fication of the atom — though there might be local eddies. The final 

 stage would be that in which all the atoms contained only one cor- 

 puscle. This result depends upon the assumption that the positive 



