General Theory of Radiation. 385 



From the phenomena of incandesence in solids and liquids, we per- 

 ceive that there exists in the incandescent bod}'' a great variety of f un- 

 lamental tones while it is in the incandescent state, because its spec- 

 trum shows a great range in the molecular vibrations, no two rates of 

 which are or can result from the same molecules or combination of mole- 

 cules. l If a piece of platinum wire be heated by a voltaic current, and 

 the effects studied in its spectrum, the first visible effect is red color, 

 then, as the temperature increases, orange is added to the spectrum, 

 then yellow, then green, blue, indigo, and lastly violet. But as each 

 of these colors is added it is not done at the expense of the first ones; 

 on the contrary, with each addition .the ones already developed are in- 

 tensified and become more vivid. That is, the amplitude of these lower 

 tones is increased by the increase of the energy, but not their rapidity 

 of vibration, because color is pitch, and the pitch and color change onty 

 by difference in wave length, and the wave length can change only as 

 the form of the body in vibration is changed. The sounds made on a 

 flute depend on the vibrations of the column of air contained in it, and 

 that column is altered in length by the fingering being shortest when 

 all the finger holes are open, and longest when they are all closed. The 

 platinum wire when heated is like seven flutes, each sounding a single- 

 note at one time or, rather, like an indefinite number of flutes pitched 

 to all the intermediate shades of tones in the octave. Therefore, al- 

 though platinum is rated as an original element incapable of being an- 

 alyzed or reduced any further, we are bound to conclude that when 

 brought to an incandescent condition it undergoes a vast number of 

 molecular rearrangements which are all maintained at once in its differ- 

 ent parts, and each of which vibrates in its own fundamental tone. 

 But more than this ; if by means of a filter we cut off all the visible 

 rays we shall find that the platinum is also radiating tones of vibration 

 in one or more octaves below the visible rays. From what has been 

 said of harmonics, we may reasonably infer that many of the tones in 

 the upper octaves are the harmonics of the octaves below, induced by 

 the increase of the energy of the temperature, in the same way as the 

 octave is raised on the flute by simply harder blowing. But it appears 

 to be an unavoidable conclusion that there are, in so-called homogene- 

 ous and elementary bodies, molecules of different forms. Heating a 

 solid body to incandescence causes the breaking up of its molecules into 

 so many forms that waves f all the lengths in the 44th octave appear 

 to be radiated. But if the body be heated till it becomes a vapor, ra} r s 

 of a great many wave lengths are suppressed, and the spectrum, which 

 before was a solid body, becomes reduced to a mere skeleton, a line 

 here and there across the spectrum only remaining. This fact must be 

 1 Draper's experiment. 



