236 Mr. J. J Thomson on the 



core is small compared with the radius of either of the rings ; 

 now, as the vortex-rings approach each other they alter in 

 size — the one in front expands and the one in the rear con- 

 tracts. If the rings are to remain together their radii must 

 become nearly equal as they approach each other, and their 

 planes become nearly coincident : it is evident, however, that 

 for this to happen the radii of the rings before they pair must 

 lie within certain limits. The energy of the gas, however, 

 and therefore the temperature, depends upon the mean radius 

 of the vortex-rings which form the atoms of the gas, and, 

 conversely, the mean radius of the vortex-rings is a function 

 of the temperature ; and if the mean radius is between certain 

 limits the temperature must also be between limits ; thus 

 unless the temperature is between certain limits the atoms 

 will not remain long together after they have paired, and so 

 chemical combination will not take place ; this reasoning 

 would indicate that chemical combination could only occur 

 between certain limits of temperature, and this seems to be 

 the case in at any rate a great many cases of chemical com- 

 bination. 



" The following reasoning will explain how it is that the 

 compound after it is formed can exist at temperatures at 

 which the element of which it is composed could not combine. 

 When the elements have once combined, the molecules of the 

 compound will settle down so that the radii of their vortex- 

 atoms will be distributed according to a definite law, and a 

 large proportion of the vortex-atoms will have their radii 

 between comparatively narrow limits, just as, in the ordinary 

 theory of gases, Maxwell's law gives the distribution of velo- 

 city. Now suppose that a molecule of a compound of the 

 elements A and B is subjected to any disturbance tending to 

 change the radii of the atoms ; though the difference in the 

 changes in the radii may be sufficient to cause the atoms to 

 separate, yet since the atoms were close together when they 

 were disturbed the difference in the changes must be small; 

 and since the motion is reversible the atom A would only 

 have to suffer a slight change to be able to combine again 

 with a vortex-ring like B, or it could combine at once with a 

 vortex-ring differing only slightly in radius from B. Thus A 

 will have plenty of chances of recombination with the B 

 atoms, and will be in a totally different position with regard 

 to them from that in which it would have been if it had not 

 previously been in combination with a B atom." 



As the question of the splitting up of the molecules is a 

 very important one, we shall proceed to consider it more in 

 detail. The configuration of two vortex-rings moving round 



