377 



spectrum of the flarne of alcohol, which are invisihle in that of the 

 flame of oil of turpentine. 



These discrepancies are shown, in the present paper, to arise from 

 the predominance of the light of incandescent solid carbon in some 

 flames, and its comparative absence in others : and it is also proved 

 that in order to obtain uniform results from the flames of the various 

 compounds of carbon and hydrogen, it is sufficient, in cases where 

 the body contains much carbon, to convert the carbon into carbonic 

 acid, without its previous separation in a solid form, by means of 

 an artificial supply of air. This is conveniently eff'ected for coal-gas 

 by means of the Bunsen gas-lamp, which burns a mixture of gas 

 and air ; and, for other bodies, by directing a stream of air from a 

 table blow-pipe through the flame. 



When thus treated, all the compounds of carbon and hydrogen 

 which have been submitted to experiment, were found to produce 

 identical spectra; that of the Bunsen lamp serving as a standard of 

 comparison. 



In these spectra five principal bright lines were observed, accom- 

 panied by several smaller ones, and separated by dark intervals. 

 One of the lines, the well known R of Fraunhofer, has been long 

 known to coincide with the line D of the solar spectrum. Two 

 other extremely close coincidences were discovered. One between 

 a brilliant green line of the lamp spectrum, and the remarkable triple 

 line h of Fraunhofer ; and another, between a bright purple line, 

 and the conspicuous line G of the solar spectrum. 



It follows, from these experiments, that all bodies whose composi- 

 tion is expressed by the general forniulse 



Cp Hr , or Cp Hr Os, 

 produce, in burning, perfectly identical spectra ; the nature of the 

 light being always the same, notwithstanding variations in the rela- 

 tive proportions of carbon and hydrogen, and the occasional presence 

 of oxygen in the body. 



