the Light of Bodies in a State of Comhujimi. \gj 
gradual manner which I have already defcribed. A piece of 
rotten wood, whilft burning, will exemplify and confirm the 
preceding illudration. When influenced by the external air 
only, if examined through a prifm, no rays will be found to 
efcape but the orange and the red. By blowing upon the 
burning wood with a pair of bellows, the corn bullion, 
being increafed, will affed thofe internal parts of the body 
which were not adted upon before. Thefe parts, therefore, 
will begin to lofe their light, and a prifm will fhew the 
green, the blue, the violet, and indigo, all appearing in fuc- 
ceffion. Appearances fimilar to the preceding may be ob- 
ferved in a common kitchen fire. When it is fainted, its 
colour is mod red, the other rays having been emitted, and 
the combuffion at a Hand ; but by blowing upon it in this 
date, its brightnefs will be increafed, and more and more of 
the rays which are yielded by the internal parts of the body 
will come to the eye, till at length, by continuing to blow, the 
combuftion will be made fo complete as to yield all the rays, 
or to make it appear perfedly white. 
Many are the varieties difcoverable in the flames and in the 
appearances of fixed burning bodies to which the preceding 
obfervations may be applied ; but, to avoid unneceflary amplifi- 
cation I will take notice only of what appears to me an imper* 
fedtian in Sir Isaac Newton’s definition of flame. He con* 
jedlures, that it may be a vapour heated red-hot. I' think 1 
fhould rather fay, that flame is an indance of eombudion 
whofe colour will be determined by the degree of decompofi- 
tion which takes place. If it be very imperfedt, the mod 
refrangible rays only will appear. If it be very perfedt, all 
the rays will appear, and- its flame will, be brilliant in 
p roper tica 
