92 Prof. Draper on the Nature of Flame, 



tlie nature of the other regional divisions distinguishable in such 

 a flame. There must be a blue portion below j blue^ because it 

 consists of the most refrangible rays, which issue forth in abun- 

 dance, for there the exterior air is most copiously and perfectly 

 applied. At the upper end of the flame, particularly if the wick 

 be long and the supply of combustible matter abundant, the 

 light emitted is red, for the products of combustion ascending 

 past that part, make it difiicult for the exterior air to get access. 



Upon these principles we may also predict what colour a 

 flame will have when we vary the circumstances of its burning. 

 Tallow or wax at temperatures greatly beneath their usually un- 

 derstood point of combustion, oxidize with a pale violet phos- 

 phorescent light, quite perceptible nevertheless in a dark room; 

 and here the light is violet, for the supply of combustible matter 

 is small, and that of the air abundant. The oxidation is there- 

 fore thorough and prompt. Eor a like reason, sulphur, as we 

 commonly see, burns blue; but if a piece thereof is thrown into 

 nitrate of potash ignited in a crucible, the light yielded is of 

 intolerable brilliancy, and absolutely white. Its whiteness does 

 not depend upon the physiological fact, that any colour, if it be 

 intensely brilliant, will seem white to the eye ; but it is optically 

 white, as is proved by prismatic examination, when all the 

 colours are perceptible. And the reason of this is, that at the 

 high temperature to which the sulphur is exposed, it volatilizes 

 faster than the nitrate of potash and air together can oxidize it, 

 and ofi'ers every intermediate rate of combustion, and emits rays 

 of every refrangibility. 



In like manner it may be shown that carbonic oxide must 

 burn with a blue flame, and cyanogen with a red. We can also 

 foresee what must be the optical result of resorting to unusual 

 methods of combustion, as when we throw into the interior of a 

 flame a jet of air from a blowpipe. In this case we destroy the 

 red and orange strata, replacing them by bluer colours. Ex- 

 amining such a blowpipe cone by the prism, we have a beautiful 

 demonstration that such has actually taken place. 



There is one of these special cases which deserves attentive 

 consideration in connexion with the appearance of the electric 

 light, — it is the production of Fraunhoferian lines when things 

 have been arranged in such a way that an incombustible material 

 is present in the substance to be burnt. This state is perfectly 

 represented in the case of cyanogen, which contains more than 

 half its weight of incombustible nitrogen. When the peach- 

 coloured nucleus of the cyanogen flame is properly examined, it 

 yields a series of dark lines and spaces exceeding in number and 

 strength those of the sunlight itself. These fixed lines are the 

 representatives of dark shells, supei-posed among the shining 

 ones with definite periodicity. la such 51 cyanpgeji flame they 



