REACTION OF PLANETS UPON SUN PUISEUX. 161 



which we wish to detect in the motion of the sun, requires years to 

 change sensibly the apparent position of a star, it takes only a moment 

 to alter the quality of its light. Whatever the distance, the light 

 waves will come to us sometimes more fi-equently, sometimes less ; their 

 path through a prism will consequently be found altered and the 

 fine metallic lines of the spectrum recorded by a photograph will be 

 displaced relatively to those of a stationary source, such as an elec- 

 tric spark used for comparison. 



The earliest happy applications of this principle were due to Hug- 

 gins and to Vogel. It was used to separate numerous double stars 

 composed of pairs of suns so close to each other and so distant from 

 us that each pair appeared as a single star. But the brightness of 

 each was sufficient to record a spectrum and the relative velocities 

 were sufficiently variable so that two spectrum lines of the same 

 chemical origin separated periodically. Subsequently another class, 

 yet greater in number, was found in which the spectrum lines were 

 not doubled, but showed a periodic oscillation. In this case we may 

 suppose that one of the two stars, while not bright enough to reg- 

 ister its spectrum, is yet heavy enough to swa}^ its associate. The 

 period is usually several weeks or days. The displacements of the 

 lines correspond to velocities of the same order as those of the 

 planets, from 10 to 100 kilometers per second. Because of the ex- 

 treme accuracy and care in the use of spectroscopes, certain as- 

 tronomers can now measure velocities to a fraction of a -kilometer. 



The time will come when pairs like the sun and Jupiter can be 

 detected, however distant they may be, provided only that the prin- 

 cipal star is bright enough to record its spectrum. Campbell, who 

 is the leader in this class of research, estimates that on the average 

 one star in three will be found spectroscopically double. It is very 

 probable that even more stars art double since we can see no reason 

 why a planet like Jupiter should be exceptional. We may predict 

 that all stellar spectra will be found thus variable even after cor- 

 recting for the orbital movement of the earth. We may then gather 

 photographic evidence of the existence of planets about the stars as 

 well as the periodic oscillation of our sun due to Jupiter. The earth 

 of course will produce a similar effect only less in amplitude and 

 period. But who would dare to put a limit to the skill of our opti- 

 cians or the patience of our astronomers in a path so definitely 

 marked out? 



THE PLANETS AS THE CALTSE OF THE SOLAR CYCLE. 



To find that we disturb the sun is of course something to elate us. 

 We will feel perhaps a more tangible satisfaction if we can find 

 that we cause changes in the aspect of its surface, disturbances visi- 

 ble by direct and not indirect evidence in the field of the microscope. 

 44S63°— SM 1913 11 



