MOLECULAR ENERGY 299 



specific heat would then be the sum of the specific 

 heats as found for each of the types of oscillators.) The 

 determination of these frequencies may be made by 

 several methods, of which two will be described. 



It is found that substances absorb the same fre- 

 quencies as they naturally emit. Thus glass which 

 transmits red light absorbs blue ; when heated it will 

 be found to give off a blue light, not a red light. The 

 frequencies which are reflected from the surface are 

 found to be those which are most strongly absorbed 

 in transmission. Thus for this case, blue light will be 

 reflected from the outer surface. 



To determine what frequencies are reflected from a 

 substance it is then only necessary to illuminate it by 

 a source containing all frequencies 1 and to observe 

 the reflected light. Now, at first thought, this does 

 not seem reasonable, for we know if we observe the 

 image of an incandescent electric lamp as reflected by 

 a sheet of red glass that the lamp appears white, 

 exactly as well as if viewed directly. In this particular 

 case there happens to be reflected from the polished 

 surface enough light of all the frequencies which com- 

 pose white light so that we do not observe the expected 

 effect. But let us arrange other mirrors of red glass 

 and see the lamp by reflection from several surfaces 

 instead of only one. Each successive reflection re- 

 sults in purer and purer light "of the natural fre- 

 quencies of the red glass. " The rays of light which 

 survive a number of such reflections are called "re- 



*A "full radiator," containing all frequencies from zero to in- 

 finity, sometimes called a "black body," since the ideal black body 

 would absorb all frequencies. 



