32 



from a to h, in such a manner that the broken light of 

 the first prism passes into the second, from the second 

 to the third and so on the light finally passing out in- 

 to an open space, where it is reflected in a diagram of the 

 seven solar colours. To this diagram, from above is 

 fitted a tube K in which the whole colour-diagram is 

 visible, and, below, a microscope is so arranged that, 

 by moving it, the slightest cross-line can be detected in 

 any part of the diagram. 



The light outside falls into the interior of this appa- 

 ratus through another special tube L\ furnished at its 

 mouth with a small prism O. The light entering 

 through this prism and tube is decomposed by all the 

 six above-mentioned internal prisms. 



In order to carry out any sort of spectroscopical ex- 

 periment an incandescent lamp L is placed a little to 

 one side of the direction of the tube L'., but in such a 

 way that the light broken by the prism should pass di- 

 rectly through the tube to the six inside prisms, and be 

 reflected in the inner chamber of the apparatus in the 

 colours of the rainbow. It is noted in these experiments 

 that, when the body under observation is heated to the 

 point of combustion, in some one place in the diagram 

 of rainbow colours, appears always a crossline of light, 

 according to the place of its appearance, peculiar only 

 to its own metal or mineral. Any other metal or mi- 

 neral on being so heated, furnishes a crossline in some 

 other place, and however many experiments we conduct 

 with different metals, each substance will throw the line 

 in a different place on the spectrum. For example glow- 

 ing gold-wire, iron, brass and platinum each give out 

 their own light-line in different places on the diagram, 

 and a similar result ensues in the case of incandescent 

 minerals. If however on the appearance of these light 

 lines we place an electric light behind the incandescent 

 lamp, the light-lines on the spectrum darken, though 

 they remain in their former places. A similar pheno- 

 menon is remarked in incandescent metals, when they 



