RADIATION 49 



structed a more accurate diagram of the same kind. We 

 have now to examine the distribution of heat in the spec- 

 trum of the electric light; and for this purpose we shall 

 employ a particular form of the thermo-electric pile, de- 

 vised by Melloni. Its face is a rectangle, which by means 

 of movable side-pieces can be rendered as narrow as de- 

 sired. We can, for example, have the face of the pile 

 the tenth, the hundredth, or even the thousandth, of 

 an inch in breadth. By means of an endless screw, 

 this linear thermo-electric pile may be moved through 

 the entire spectrum, from the violet to the red, the 

 amount of heat falling upon the pile at every point of 

 its march being declared by a magnetic needle associated 

 with the pile. 



When this instrument is brought up to the violet end 

 of the spectrum of the electric light, the heat is found to 

 be insensible. As the pile is gradually moved from the 

 violet end toward the red, heat soon manifests itself, aug- 

 menting as we approach the red. Of all the colors of the 

 visible spectrum the red possesses the highest heating 

 power. On pushing the pile into the dark region be- 

 yond the red, the heat, instead of vanishing, rises sud- 

 denly and enormously in intensity, until at some dis- 

 tance beyond the red it attains a maximum. Moving 

 the pile still forward, the thermal power falls somewhat 

 more rapidly than it rose. It then gradually shades 

 away, but, for a distance beyond the red greater than 

 the length of the whole visible spectrum, signs of heat 

 may be detected. 



Drawing a datum line, and erecting along it perpen- 

 diculars, proportional in length to the thermal intensity at 

 the respective points, we obtain the extraordinary curve, 



SciENOiJ — —-3 



