Xybfervations on inflammable Air. *59 
which may be eafily diftinguithed between the body o r 
the flame by their vivid light. Thefe fparks, which are 
of a vivid colour, dart in every direction. They might 
be eafily taken for thofe fparks that are emitted from 
red-hot iron ; or they might be compared to very final 1 
grains of gunpowder, if thefe were inflamed fuccefiively, 
and without fmoke ; or they might even be compared to 
charcoal that fparkles, but without any noife. This phe- 
nomenon feems very interefting, as it refpedts the nature 
of the inflammable air itfelf. What feems to me moft 
Angular is, that this appearance forms a diftinCtive cha- 
racter between the inflammable air of metals, and that 
extracted from animal or vegetable fubftances ; at leaft I 
may fafely fay, that I never found the inflammable air 
"of animal or vegetable fubftances fparkle like that ex- 
tracted from metals. In feveral of the former kinds of 
air I could obferve no fparkling at all; in others the 
fparks were fo few that they might be confidered as no- 
thing in comparifon to the fparkling of the inflammable 
air from metals. 
The inflammable air of metals itfelf, if left in contact 
with water for a long time, or fliook in it till it becomes 
lefs inflammable, will in great meafure lofe its fparkling 
property, and at laft lofes it intirely, when it is become 
in a ft ate of being hardly inflammable. I have obferved, 
B b b 2 that 
