340 NINETEENTH CENTURY. FT. HL 



of the slit was lighted by the sun and the other half by a 

 luminous gas. In this way No. 2, Plate L, would appear 

 above, and No. 3, for example, immediately below it. 



WhPc doing this they could not help remarking that the 

 bngfit yellow line of the sodium spectrum, No. 3, was exactly 

 in the same position as the black line, D, in the solar spec- 

 trum ;' and KirchhorT found that when he passed a faint ray 

 of sunlight through luminous sodium vapour (so as to make 

 the two spectra, 2 and 3, cover each other), the yellow line 

 exactly filled the black line with its light. He now wished 

 to see how bright he could make. the solar spectrum without 

 overpowering the light of the sodium, so he let the full sun- 

 shine pass through the sodium flame. To his great aston- 

 ishment he saw the black line at D start out more strongly 

 than ever. The sodium light while being overpowered itself 

 had absorbed some of the yellow light of tJic sun ! 



This suggested to him the idea that the black line D must 

 be caused by the white light from the sun passing through 

 sodium vapour before it reaches us. There was a very simple 

 way of proving whether this were so, for burning solids, you 

 remember, give a continuous spectrum (i, Plate I.); there- 

 fore, if he could produce a dark line by passing the light of 

 a burning solid through glowing sodium vapour, he would 

 imitate one of the defects in sunlight. So he burned a lime- 

 light, and when he had the continuous coloured band in his 

 spectroscope, he burned a flame coloured by sodium between 

 the lime-light and the prism. The experiment was quite 

 successful ; the dark space, D, started out upon the spectrum, 

 and thus he proved beyond doubt that incandescent or glow- 

 ing sodium vapour absorbs out of white light exactly those same 

 rays which it gives out itself when glowing. 



1 This line is really formed of many lines which cnn he seen in a 

 powerful spectroscope, but it appears single in a small instrument. 



