36 
COMBUSTIBILITY OF GASES. 
Repetition of 12. In order to ascertain more clearly the nature of the gas, 
variations! ^ tu ^e with mercury, and then introduced a drop of 
alcohol. I heated the alcohol till the mercury had descended 
one inch below the metallic cylinder, which served to conduct 
the electric spark. I brought this small apparatus near the 
conductor of the electric machine, and passed several hundred 
electric sparks through it, at the same time that I was careful 
to keep up the dilatation by heating the tube from time to time. 
The electric light, under these circumstances, shewed a beauti- 
ful green colour, and formed a torrent of celadon green light, 
which bad a beautiful effect in the dark. 
Dccompos. of While the tube became cold, the alcoholic vapor was con- 
hol b^eiectri" ^ enset ^ * ntoa thin stratum of liquid, which, by the rise of the 
city. mercury, approached the metallic cylinder. At the moment the 
stratum of alcohol touched the metallic cylinder, a decompo- 
sition of the alcohol, in the liquid state, commenced. The 
surface of the mercury then became covered with bubbles 
scarcely visible, which were from time to time inflamed by the 
electric sparks, and formed in the middle of the liquid small 
brilliant points, resembling the flame of phosphorus. It ap- 
pears that alcohol is slowly decomposed in oxigen, and in in- 
flammable gas. The alcohol in vapor is still more rapidly de- 
composed ; but the product is an uniform gas, composed of 
oxigen, hydrogen, and carbon. By a quarter of an hour’s elec- 
tric action, I obtained half an inch of this gas, which is car* 
bonated hydrogen, as I have proved by experiment. 
I could not decompose the vapor of water in so short a time 
by means of the electric spark. It appears that water is more 
difficult of decomposition, or that its elements reunite slowly 
after their separation to recompose water. 
Green colour 13. q^e beautiful celadon green, which the electric spark 
of the electric . . * 
spark in vapor assumes in the vapor of alcohol, induced me to examine seve- 
ef alcohol. ra ] other bodies in the state of gas. According to Priestley, 
the electric spark is of a purple red in hydrogen gas j but as 
it takes this tint in every other dilated gas, (as, for example, in 
air dilated by the air-pump,) we might, in the former case, 
attribute the colour to the rarity of the hydrogen. The nature 
of the electric fluid must nevertheless have some influence on 
the colour of the electric spark $ for the spark appears red in 
am moniacal gas, and in phosphorated hydrogen, though those 
fluids 
