172 ANNUAL REPORT SMITHSONIAN INSTITUTION, 192 8 



sands of years. Where, then, shall he find a sonree of energy to 

 last niiilions of millions of years ^ 



More than 20 years ago I directed attention to the enormous 

 store of energy made available by the annihilation of matter, by 

 positive and negative electrons falling into and annihilating one 

 another, thus setting free the whole of their intrinsic energy as 

 radiation. On this scheme neither energy nor matter had a perma- 

 nent existence, but only a sort of sum of the two ; each was, the- 

 oretically at least, convertible into the other. ^Vhether energy [s 

 ever transformed into matter we do not know; probably not. But 

 the falling together of electrons and protons forms the obvious 

 mechanism for the transformation of matter into energy, and it 

 now seems practically certain that this is the actual source of the 

 radiation of the stars. A beam of radiation exerts pressure on 

 any surface it falls upon, just as a jet of water does or a blast of 

 air. The reason is that radiation carries mass about with it, and 

 electromagnetic theoiy tells us the amount of this mass. For ex- 

 ample, we can calculate that a searchlight which is radiating 50 

 horsepower of energy is discharging mass into space with the radia. 

 tion at the rate of a gram and quarter a century; with sufficiently 

 delicate adjustments it might even be possible to observe the recoil 

 of the searchlight. Indeed, the pressure of radiation has actuall}^ 

 been measured although not in this particular way. New mass is of 

 course being continually fed into the searchlight by the electric 

 current. 



Each square inch of the sun's surface is in effect a searchlight 

 discharging radiation into space at the rate of 50 horsepower, and 

 so is discharging mass at the rate of a gram and a quarter a century, 

 and the sun's surface is so large that the sun as a whole is discharg- 

 ing mass into space at the rate of 250 million tons a minute. Now 

 the sun has no source of replenishment. It must have weighed 

 360,000 million tons more yesterday than to-day, and by to-morrow 

 will weigh 360,000 million tons less. These are not mere speculative 

 statements; they rest on observation, and on generally accepted 

 principles which are directly confirmed by observation. 



Allowing for the fact that a more massive star emits more radiation 

 than a less massive one, we can calculate that 5 or 10 million million 

 years ago the sun must have been several times as massive as it is 

 to-day, so that it has already lost most of the mass it had at birth. 

 Of each ton it had at birth only a few hundredweights at most i-o- 

 main to-day. The loss of mass which accompanies radiation is, 

 then, no mere academic hairsplitting. It is a real astronomical 

 phenomenon, and young stars must be many times as massive as 

 old stars. 



