COLOUR OF FLAME. 17 



atmospheric air, I observed very particularly whenever an 

 undue increase of pressure was put on for the moment, experi- 

 mentally, that there were exhibited the distinct colors of red, 

 bluish, purple, and yellowish light, forming separate portions of 

 the volume of flame, but that when the pressure was adjusted, 

 as is usual, and moderated to a regular proportionate supply, 

 an excellent intense white light was produced. 



Such diversity of color in the flame occasioned by the excess 

 of velocity of the stream of atmospheric air, evinced the peculiar 

 color of the oxygen red, of the nitrogen yellow, and of the 

 carbon, a composition of the three original elements, produced 

 from the residuum of the inflammable bodies submitted to 

 ignition, which when first ignited burned with a dull and 

 deficient light, full of intense smoke, which latter was imme- 

 diately dissipated, and the flame expanded and projected four 

 feet high, affording a most brilliant clear white light from the 

 action of a very moderate stream of atmospheric air. 



As a proof of the identity of the red ray with positive elec- 

 tricity, Mr. J. H. Abraham, F. L. S., of Sheffield, informed 

 me that in trying some experiments on lateral discharges, and 

 on the gradual and silent discharge of a battery or jar over- 

 flowing with positive electricity, he obtained sparks seven 

 inches long of a beautiful vivid rose-colored light, while I have 

 observed from negative electricity the most beautiful light- 

 green color from a galvanic arrangement operating to produce 

 a magnetic rotary power. 



The accordance of these facts would show first that oxygen, 

 the red ray, positive electricity, and positive galvanic power, 

 with inflammation, are IDENTICAL. 



And that nitrogen and hydrogen, the yellow and blue rays, 

 affording magnetic power, the negative electricity and galvanic 

 negative essence are also identical. 



That by their union they form light, and by their accumula- 

 tion to the surfaces in the pervasion of masses of matter, they 

 may (and in my opinion do) by their opposite states produce 



c 



