PHYSICS-NINETEENTH CENT.-SPECTROSCOPY. 483 



vapours of nitrous acid had a remarkable effect on the spectrum. 

 When the sun's light was passed in its course to the prism through a 

 glass vessel containing these vapours, the spectrum was seen to be 

 crossed by a great number of dark bands, which were quite inde- 

 pendent of the Fraunhofer lines, for as was afterwards shown by W. H. 

 Miller and Professor Daniell, they were also produced with other 

 sources of light ; for instance, with the bright part of a candle-flame, 

 which, but for the interposition of the nitrous acid vapours, would have 

 given a continuous spectrum. By the continuous spectrum we must under- 

 stand one without any lines, either dark or bright, but showing the 

 various prismatic colours graduating one into the other. The appear- 

 ance presented by the spectrum of these nitrous acid vapours is re- 



FlG. 221. 



presented in Fig. 221. In the course of his investigations Brewster 

 observed that at times certain dark lines were visible in this spectrum 

 which on other occasions ceased to appear. He soon found that these 

 lines appeared when the observation was made with the sun low or 

 near the horizon, and that they ceased to be visible when the sun had 

 attained a certain elevation. He was therefore led to attribute them 

 to the action of our atmosphere upon the solar rays, an action which 

 appeared to bear the closest analogy with that which was exercised 

 by the brown-red vapour of the nitrous acid, namely, a power of 

 stopping or absorbing certain of the solar rays. The dark lines which 

 result from this property of gases or vapour were henceforth known 

 as absorption bands or absorption lines, and we shall presently see the 

 important part they play in the development of the science. 



Brewster's experiments were followed up by others instituted by 

 W. H. Miller and Daniell, who very shortly afterwards showed that 

 other coloured vapours also give absorption bands. They ascertained 

 that this was the case with the vapours of bromine and iodine, and 

 with a certain yellow-coloured gaseous compound of chlorine and 

 oxygen. The lines, however, given by each substance are entirely 

 different. They appear to have no relation to the colour of the gas 

 or vapour. Thus bromine and nitrous acid vapours, although they 

 have almost identically the same colour, give quite different sets of 

 absorption bands in their spectra ; and again, the red vapours of chlo- 



31 2 



