TIIE CAUSE OE THE LUMINOSITY OF FLAME. 
129 
the increased temperature; but we found that the increase of 3588° F.,—the dif¬ 
ference between the temperature of hydrogen burning in air and burning in oxygen,— 
did not increase the light. I don’t say that this is conclusive, but it is remarkable, and 
it is not a solitary case. If we take hydrogen and burn it in chlorine, we get a little 
more light than when we burn it in atmospheric air ; but still the luminosity is very 
feeble. Here is a vessel containing a mixture of hydrogen and chlorine, which has been 
excluded from the light of day, because the action of light induces the combination of 
these two gases. If we explode this mixture in soap-bubbles, it produces the same in¬ 
tense noise as the mixture of oxygen and hydrogen, without giving much more light. 
But if we burn it in a close vessel, where it cannot expand, the light is intense. Again, 
on burning carbonic oxide and oxygen, under common atmospheric pressure, though we 
get more light than with hydrogen and oxygen, the increase is but trifling. Burning 
carbonic oxide in the air, we get, according to Favre and Silbermann, a temperature of 
5122°; burning it in pure oxygen, the temperature is increased by 7072°, that is, to 
12,794° ; but the increase of light is very little. Burning the same mixture in a closed 
vessel, we have a still more brilliant light than before. In all these cases we obtain a 
powerful luminous effect when the gases are not allowed to expand, although in none 
of them is there a single particle of solid matter present. [All these results were expe- 
mentally obtained.] 
“Many other instances might be adduced of the production of light where there are 
no solid particles; and the following additional cases will serve to clear the way to the 
explanation of the true source of light in the gas-flame. The combustion of metallic 
arsenic in a stream of oxygen,—effected by placing a fragment of the metal in the bulb 
of a glass tube, through which the oxygen passes (a hood being lowered over it to carry 
off the fumes),—gives a great amount of light, though no solid particles are present. 
Again, metallic arsenic plunged into a hydrogen gas-flame volatilizes, and the vapour 
produces a decided luminosity, but inferior to that produced in the stream of oxygen. 
If, instead of metallic arsenic, we take a piece of arsenious acid,—the product of the 
combustion of metallic arsenic,—and try the same experiment upon it, the effects are 
similar to those produced with the metal. A compound of arsenic has, in fact, beeu 
employed in this way as a signal light. *It is the basis of the substance called “ Indian 
fire,” which has sometimes been used for signalling in trigonometrical surveys. By 
burning this “ Indian fire,” we have, virtually, the combustion of arsenic and sulphur in 
oxygen ; we get a very intense light, and yet there is probably no substance in a solid 
condition in the flame. 
“ Here, then, are many facts and considerations which shake our faith in the theory 
that the light in the flame of coal-gas is due to the presence of solid particles. In the 
next lecture we will resume this subject, and will endeavour to decide what is the true 
source of light in a gas- flame. 
“ In the last lecture we glanced at the apparatus and processes employed in the ma¬ 
nufacture of coal-gas, and then proceeded to investigate the conditions of luminous 
combustion. Our experiments led us to doubt the proposition that the luminosity of 
the gas-flame is due to the incandescence of solid particles of carbon. We found that 
the combustion of the so-called non-luminous gases under certain conditions is attended 
with a very brilliant light. By igniting mixtures of hydrogen and oxygen, carbonic 
oxide and oxygen, and hydrogen and chlorine in close vessels, so that their expansion 
was prevented, we got striking luminous effects; yet we know that the flames produced 
by the combination of these gases cannot owe their luminosity to solid or liquid par¬ 
ticles. The combustion of metallic arsenic in oxygen furnished another instance of the 
production of intense light by flame free from solid matter. Let us push this subject a 
stage further before we leave it. 
“ First, we will burn some of the liquid called bisulphide of carbon in contact with 
air. [The liquid was poured into a dish, and ignited beneath the glass hood.] This 
compound burns with a lambent blue flame, about as luminous as that of carbonic 
oxide. The flame is large, but the light emitted by it is extremely feeble. Though 
carbon is a constituent of the fluid, when I expose this plate of white porcelain to the 
flame I do not get a trace of any black deposit. Looking closely at the plate, I can 
detect a slight deposit of sulphur, the other constituent of the liquid; but as sulphur 
assumes the gaseous condition far below a red heat, no solid particles of this body can 
exist in the flame. By burning this bisulphide of carbon in oxygen a much higher 
VOL. IX. K 
