■the Ll^ht of Bodies in a State of Comhufic-tu 1 93 



npon any hard globular fubftance, Gradually bring the paper, 

 thus formed, to that diftance from the candle at which it will be- 

 gin to take fire. In this cafe a beautiful blue flame may be feen, 

 hanging as it were by the paper till a hole is made in it, when 

 the flame, owing to the increafed a6lion of the air upon all. 

 parts of it, becomes white, though the edges ftill continue of 

 a blue or violet colour. As a confirmation of what I have 

 concluded from the preceding fa£ls, it may be obferved, that 

 the very flame which, when expofed to a certain degree of 

 heat, emitted the mofl refrangible rays only, will, if expofed 

 to a greater degree of heat, emit fuch as are lefs refrangible* 

 The flames of fulphur, fpirits of wine, &c. when fuddenly 

 expofed to the heat of a reverberatory, change their blue ap- 

 pearance for that .which is perfe6lly white. But to gain a 

 more flriking diverfity of this faft, I adopted Mr. Melvill*s 

 mode of examining bodies whilfl: on fire. I darkened my 

 room, and placed between my eye and the combuftible a (lieet 

 of pafteboard, in the center of which I made a fmall perfora- 

 tion. As the light of the burning body efcaped through this 

 perforation, I examined it with a prifm, and obferved the fol- 

 lowing appearances. When the fpirits of wine were fet on 

 fire, all the rays appeared in the perforation ; but the violet, 

 the blue, and the green, in the greatefl abundance. When 

 the combuflion of the fpirits was checked by throwing fome 

 fal ammoniac into the mixture, the red rays difappeared ; but 

 when, by the long continuance of the flame, the 'fal ammoniac 

 was rendered fo hot as to increafe, rather than diralnifh thci 

 combuflion, the red rays again appeared at the perforation. If 

 the fcreen was managed fo that the different parts of the 

 fiame might be examined feparately, I always obferved that 

 Vol. LXXV. C c - the 



