1 4 Drs Christison and Turner on the Construction of 
As to the limit at which the light ceases to increase with the diminution 
of the supply of air, we have invariably remarked, that nothing farther is 
gained, after the flame begins to be tipped with brown. According as the 
aperture is diminished, the flame loses its white colour, and gradually acquires 
a yellow, and at length a brown tint. At first a great increase of light is 
gained with very little sacrifice of the purity of the tint. In the second ex- 
periment mentioned above, the damped flame (if we may use the expression) 
of 3| inches seemed almost as purely white as the original 2-inch flame. But, 
beyond that point its lustre was considerably impaired, although its light was 
still manifestly increased. Whenever it began to be tipped with brown, any 
farther diminution of the air also diminished the light. 
These facts lead to the conclusion, that, in order to obtain the greatest possible 
light with any burner, the supply of air ought to be such as to burn the gas com . 
pletely ; but that, when it is burnt completely, nothing is gained, on the con- 
trary, much is lost, by supplying more air to render the combustion more vivid. 
This is the principle to which we have several times alluded, as being the 
guide we have followed, while endeavouring to determine the various points 
in the construction of the burners. The combustion of the gas should not be 
more vivid, than is sufficient to render it complete. 
The cause of the loss of light sustained by too free a supply of air will be 
found, we apprehend, in the clear and ingenious explanation given by Sir 
Humphrey Davy, of the source of the light of the gases. Sir Humphrey 
supposes, that a white light is given out only by those gases which contain an 
element of so fixed a nature, as not to be volatilizable by the heat caused 
during the combustion of the gas ; and that in coal-gas this fixed element is 
charcoal, formed by the gas undergoing decomposition before it is burnt. The 
white light is caused by the charcoal passing into a state, first of ignition, and 
then of combustion. Consequently, no white light can be produced by coal or 
oil gas, without previous decomposition of the gas. 
That the gas undergoes decomposition before it burns, and that the carbo- 
naceous matter is burnt in the white part of the flame in the form of charcoal, 
is shown by placing a piece of wire-gauze horizontally across the white part of 
the flame, when a large quantity of charcoal will be seen to escape from it un- 
burnt. And, that this previous change is necessary to the production of a 
brilliant white light, will appear, if we consider the kind of flame which is 
produced when decomposition does not previously take place. For example, 
if the gauze be brought down into the blue part, which always forms the base 
of the flame, no charcoal will be found to escape. Or, if the gauze be held at 
some distance above the burner, and the gas be kindled not below but above 
it, by which arrangement the air and the gas are well mixed previous to com- 
bustion, the flame is blue, and gives hardly any light. The reason is obviously, 
that, in both cases, the air is at once supplied in such quantity in proportion 
to the gas, that the first effect of the heat is to burn the gas, not to decompose 
it. (On the Safety-Lamp , p. 48. et seq). 
We must refer to the author’s paper for the proofs adduced in support of 
his doctrine. 
A single step farther in the investigation of these curious phenomena, 
would have led Sir Humphrey to remark a curious fact soon to be mentioned, 
3 
