162 
3. On the Constitution of Flame. By Mr Swan. 
In this communication the author discusses the theory of the con- 
stitution of flame advanced by Professor Draper of New York, in 
his paper “ On the Production of Light by Chemical Action,” which 
appeared in the Philosophical Magazine for 1848. 
Professor Draper refers to experiments which prove that the 
higher the temperature of an incandescent body, the more refrangible 
are the rays of light emitted by it. Pie assumes that the tempera- 
ture of the outer portions of a flame is greater than that of the in- 
terior regions ; for outside there is a better supply of oxygen, and 
hence a more intense combustion is maintained. He thence argues 
that a flame must consist of a series of layers of different colours 
following the order of tints in the prismatic spectrum, the red being 
innermost and the violet outermost. Professor Draper conceives 
that he has demonstrated by experiment that such a structure 
actually exists in flame. 
Mr Swan, on a careful examination of various flames conducted 
essentially according to Professor Draper’s methods of observation, 
with the exception of one particular in which that observer’s process 
seemed objectionable, completely failed to verify his results ; and 
on the following grounds believes his views regarding the constitution 
of flame to be erroneous : — • 
ls£, Professor Draper’s method of observation, so far as it can 
be gathered from his somewhat imperfect account of it, involves an 
error in principle calculated to lead to fallacious results. 
2d, His theory is founded on the quite gratuitous assumption of 
a great diversity of temperature between portions of flame so closely 
contiguous, as to render the existence of such diversity of tempera- 
ture highly improbable. 
3d, Even although great diversity of temperature did exist in 
different portions of a flame, there is no reason to believe that it 
would give rise to a series of layers of different colours. As the 
temperature of an incandescent body is raised, rays of continually 
higher refrangibility are, doubtless, emitted ; but these are in every 
case accompanied by rays of low refrangibility. From this it follows 
that the outer regions of a flame, however high their temperature, 
will not yield exclusively the extreme violet rays of the spectrum, as 
Professor Draper supposes, but will equally emit the extreme red 
