M. A. Wiillner on the Absorption of Light. 45 



same quality, inasmuch as the vibrations of this kind of light 

 must be enfeebled and diminish in velocity, because at each 

 vibration the aether-molecules clash with the adjacent sodium- 

 molecules vibrating in the same phase. 



It readily suggested itself, with reference to the theory of ab- 

 sorption developed by Mr. Stoke3*in an earlier paper, to extend 

 this idea to all phenomena of absorption — to see the absorption 

 of light based on the fact that the vibratory molecules of aether 

 communicate their velocity to the molecules of the absorbing 

 body, which perhaps are only for the first time caused by the 

 light to vibrate in phases appropriate to their molecular struc- 

 ture. It followed further therefrom, that if any body or gas 

 at ordinary temperatures exhibited a special power of absorp- 

 tion for certain kinds of light, then the body in question, in 

 virtue of its molecular structure, is especially fitted for the cor- 

 responding oscillations. If therefore the body in question does 

 not change its nature at high temperatures, it must, if heated to 

 that temperature at which it can emit light, send out especially 

 that light which it absorbs at lower temperatures. 



To prove this theorem, and therewith to find eventually a 

 proof of the accuracy of the new theory of absorption, I endea- 

 voured last year to establish a comparison of the absorption 

 spectra of iodine, bromine, and hyponitric acid, with the spectra 

 of the flames in which these vapours are incandescent. To ob- 

 tain flames which only emit light arising from these vapours, I 

 passed the vapours into the flame of hydrogen. To bring iodine, 

 for example, into the flame, I passed hydrogen from a spacious 

 gas-holder into a tube containing iodine vapour, the glass tube 

 ending in a tolerably wide point, directed upwards, which was 

 just in front of the slit of the spectrum apparatus. In the glass 

 tube was a plug of asbestos, to prevent solid particles being car- 

 ried into the flame. Further, the front end of the tube was laid 

 in a sand-bath, and thus kept at a temperature at which iodine 

 vapour does not condense. The iodine was placed at that end 

 of the tube near which the hydrogen entered, and was heated 

 with a spirit-lamp. The hydrogen thus became completely 

 mixed with iodine vapour, which became incandescent in the 

 flame when the gas at the point was lighted. 



I operated similarly to bring vapour of bromine or of hypo- 

 nitric acid into the flame. 



I succeeded then in proving that the conclusions of the ab- 

 sorption theory were confirmed, in so far that those parts of the 

 spectra of the flames were the brightest in which most dark lines 

 appeared when daylight passed through the vapours. For 

 example, in the absorption-spectrum of iodine, dark lines predo- 

 * His investigation on Fluorescence. 



