154 SEE— DYNAMICAL THEORY [April 19, 



to the eye against the liackground of the sky, because it is condensed 

 into a beam, the same thing- obviously could develop also for par- 

 ticles in the solar corona itself, even if they be not sufficiently con- 

 centrated to present at night the aspect of a ray extending from the 

 sun. In fact such rays of charged matter are proved to emanate 

 from the sun by Maunder's researches on the sun spots and magnetic 

 disturbances noted at Greenwich, and published in the Monthly No- 

 tices of the Royal Astronomical Society for 1905. 



4. The emission of charged particles from the sun being thus 

 clearly proved, the only question remaining open to discussion is 

 whether any of the matter thus driven away from the sun goes away 

 to the other fixed stars. But as my calculations show this to occur 

 for the particles of the tails of comets which graze the sun's disc in 

 perihelion— the only case in which the beams can be distinctly seen 

 and the velocity of the particles determined from the lack of curva- 

 ture in the tails — it must, by similarity of causes and effects, be held 

 to occur also for some of the particles in the corona, even though 

 they be invisible, owing to the diffuseness of the streamers. 



5. The sun therefore is losing matter incessantly as well as gain- 

 ing it, in the form of meteorites from celestial space. And in my 

 "Researches," Vol. II., 1910, I have shown that the secular accelera- 

 tion of the earth's motion indicates that at present the gain exceeds 

 the loss; but if the sun was hotter in past ages, the reverse tendency 

 formerly may have been at work. 



6. Thus it appears to be demonstrated, by observed phenomena 

 in our planetary system, that the sun is both gaining and losing 

 matter, but that at present the rate of gain exceeds that of loss, so 

 that there is a secular acceleration of the planets of such excessively 

 minute character that it long escaped detection. In other fixed stars, 

 it is probable that various combinations of gain and loss are at work ; 

 and we may be sure that the masses of the stars are not strictly con- 

 stant over long ages, however approximately an even balance of gain 

 and loss may hold for shorter intervals of time. 



The view held by Newton and adopted by Lagrange and Laplace 

 that -the sun's mass may be considered constant, is only approxi- 

 mately true, and cannot properly be applied to the secular equations 



