MEMOIR II.] SPECTRUM ANALYSIS OF FLAMES. ^\ 



the frequency of vibration still increases, the temperature 

 correspondingly rises, and the light successively becomes 

 orange, yellow, green, blue, etc., and this condition ob- 

 tains in the successive strata of a flame, as we pass from 

 its interior to its exterior surface. 



The general principle at which I thus arrive, as the 

 result of this experimental investigation, viz., that there 

 is a connection between the energy with which chemical 

 affinity is satisfied and the refrangibility of the resulting 

 light, assumes the position of a simple consequence of 

 the undulatory theory. Is it not very natural, if all 

 chemical changes are attended by vibratory motions in 

 the particles of the bodies engaged, that those vibrations 

 should increase in frequency as the action becomes more 

 violent ? But an increased frequency of vibration is the 

 same thing as an increased refraugibility. 



VIII. On the physical cause of Fraunliofer* s dark lines. 



Although I have extended this memoir to so great a 

 length, I have omitted many facts which have been 

 made the subject of experiment. I cannot conclude, 

 however, without offering some remarks on the artificial 

 production and cause of Fraunhofer's fixed lines. 



It has been stated that I was led to expect the pro- 

 duction of these lines in the flame of cyanogen, from 

 considering the circumstances under which its combus- 

 tion takes place. Returning to this phenomenon, I 

 shall here point out a very remarkable numerical re- 

 lation existing among the fixed lines of the solar spec- 

 trum. 



The following table contains Fraunhofer's determina- 

 tion of the wave-lengths of the seven great fixed lines 

 of the spectrum, which are designated by the capital 

 letters of the alphabet from B to H. I have added the 

 wave-length of A from my own experiments. 



