9-2 
135 
122-124] The Chemistry of the Stars 
weights, those elements whose atomic weights are highest having the greatest 
capacity for the spontaneous generation of radiation by annihilating them 
selves, and, in consequence, having the shortest lives. Such elements will 
necessarily be the first to disappear as the star ages, their disappearance 
reducing not only the mean atomic weight in the star but also the mean rate 
of radiation per unit mass, since these heavy elements are the most energetic 
radiators. Just as, on the coast, the hardest rocks survive for longest the 
disintegrating action of the sea, so in a star the lightest elements survive for 
longest the disintegrating action of time, with the result that ultimately the 
star contains only the lightest elements of all and so has lost all radiating 
power. Our terrestrial elements have so little capacity for spontaneous trans 
formation that they may properly be described as “permanent.” The result 
of our previous calculations of § 104 may be stated in the form that if the 
terrestrial elements underwent any appreciable transformation in periods 
comparable with a period of 10 17 years, the resulting generation of heat by 
the earth’s mass would make the earth too hot for human habitation. Again 
the radioactive elements must be mentioned as an exception; they probably 
represent the last surviving vestiges of more vigorous primeval matter, thus 
forming a bridge between the inert permanent elements and the heavier and 
shorter-lived elements of the stars. 
The half-period of uranium (5 x 10 9 years) is so short that we must 
suppose that the supply of uranium in the sun is being continually re 
plenished. Otherwise, as Lindemann* has remarked, an interval of 5 x 10 11 
years would reduce the amount of uranium in the sun by a factor of 10 _so . 
Even if the sun had consisted wholly of uranium at the beginning of this 
period there would be less than 2 kilograms left at the end, and more than 
this exists now on the earth alone. 
It is hardly likely that this replenishment can occur through the synthesis 
of lighter elements. We have seen that a temperature of the order of a 
thousand million degrees is necessary to effect the disintegration of uranium 
and a similar temperature would, in all probability, be needed to produce 
uranium by the synthesis of its components. 
Thus we are led to suppose that the sun’s store of uranium is continually 
being formed by the disintegration of a parent element of higher atomic weight 
and far longer half-period. The amount of uranium would in this case be 
kept constant if the amounts of the two substances were proportional to their 
half-periods, and a half-period of the order of 10 13 years for the parent element 
removes all difficulties associated with the time-scale without calling for any 
improbably large amount of this element. 
The Process of Annihilation. 
124. We are now in a position to form some sort of a picture of the 
mechanism by which a star generates and radiates its energy. 
* Nature, cxv. (1925), p. 229.