THE MOVEMENTS OF THE STARS. 287 



the hydrogen is heated to incandescence. We are to sup- 

 pose that one of the modes of vibration in the molecule of 

 the gas is so timed, that the oscillations it imparts to the 

 ether are those competent to produce light of the character 

 indicated by F. Let us now suppose that a light is radiated 

 through a mass of cold hydrogen. The majority of the wave 

 lengths of the light are such that no mode of vibration in 

 the molecule corresponds to them, and consequently they 

 are allowed to pass. Those particular rays, however, 

 which vibrate with the period belonging to F, find the 

 molecules in tune with them, so to speak ; the molecules 

 accordingly commence to vibrate, and thus absorb the 

 particular energy belonging to that particular wave length. 



Let me illustrate the matter from other branches of 

 science. There are many analogous phenomena in the 

 theory of sound ; open a piano for instance, each string 

 corresponds to a particular note. Sound a note from some 

 other instrument, such as the human voice close by, and 

 the piano resounds, but this resounding is not an indis- 

 criminate return of the sound like an echo, for each note in 

 the piano remains silent unless it has been appealed to by 

 a precisely similar note from the other instrument. Thus 

 the vibrations caused by the other instrument pass with- 

 out effect over those notes with which they are not in 

 tune, while the energy that they possess is absorbed by 

 the strings which resound to them. 



We may pursue this illustration a little farther. Let 

 us suppose a multitude of piano wires stretched from 

 floor to ceiling of an apartment, all tuned to one par- 

 ticular note; now of course if all these strings were 

 struck, and if suitable sounding boards could be pro- 



