the radiant energy absorbed performs work upon the ether, presumably the generation of minute ethereal vortex-rings—the elementary particles from which electrons are derived, or possibly the positive and negative electrons themselves out of which the atoms are formed. From associations of electrons to atoms, from atoms to molecules, from molecules to the first tiny beginnings of a cosmical crystalline sublimate, there is a continual progression and increase of size. Finally, this widely dispersed material must be gathered from the immense voids of space into the germs of future worlds, and for this task the meteorites appear to be the appointed instruments.
A process which goes on forever in one direction is inconceivable. For every swing of the pendulum there must be a counter swing. If atoms have been built up by the action of light, they can be torn apart, and the energy of their formation will be once more set free. We may assume that a certain proportion of the atoms disintegrate, a very minute proportion ordinarily in planetary bodies, but a much larger one under solar conditions. The following facts suggest a relation: (1) The known radioactive elements disintegrate with the production of helium, and the evolution of enormous thermal energy. (2) The stars which are, at least externally, the hottest, since they have effective temperatures which have been rated in a few cases as high as 40,000° C, are surrounded by extensive atmospheres of helium. These relations favor the hypothesis that the helium stars contain an exceptional amount of peculiarly unstable elements, and owe their high temperature to the heat set free in the gradual elimination and destruction of these substances. The energy of formation of the atoms is being slowly dissipated as radiation from the stars, but is eventually reabsorbed by the ether, and is thus restored to the material phase of its existence by the formation of new atoms.
A plausible inference may be formed from the behavior of radium. In 1,000 years, 4 grams of radium will have been nearly one third transformed into other forms of matter of less intrinsic energy, the radium being reduced to about 2.8 grams. During this interval of. time, the four grams of radium will have emitted, according to Rutherford's measurement of the annual production of heat from radium,
gram-calories of heat. This is, of course, only a first approximation. The progression is not strictly linear. Since the gram of substance transformed has not, in this case, been annihilated as matter, but has simply been transmuted into other forms of matter, the gram-calories of thermal energy do not represent the total mass-energy of the gram of matter, but only that portion of the mass-energy which has been lost in this partial transformation. If we suppose that the total original energy is 1,000 times as much as that which has been lost
