322 On the Cause of Explosions in Lamps. 
flame shall be cool instead of hot. By these means, and 
especially by demanding that all mineral oils shall comply 
with the letter instead of what is said to be the spirit of the 
Petroleum Act of 1862, explosions in lamps will seldom or 
never be heard of. 
A few words in conclusion. I have pointed out the cause 
and the remedy. With whom will begin the adoption of the 
- means of prevention of these explosions? It is useless to 
say the refiner ought to do this; the refiner at New York 
and Philadelphia satisfies the requirements of the shipper ; 
the shipper sends whatever is wanted by the British mer- 
chant; the merchant imports according to the demands of 
the dealer; the dealer supplies the wants of the retailer; 
and the retailer finds ready sale for any mineral oil that 
will give a good light without smoke, and is cheap. Then 
the adoption of preventive measures scarcely lies with the 
public; for the consumer seldom knows anything about 
thermometers, petroleometers, igniting-points and explod- 
ing-points, and the only test I can suggest to him, by which 
to assure himself that a given specimen of oil is, or is not 
dangerous, is to unscrew the brasswork after the lamp has 
been burning for an hour or two, and to introduce the flame 
of a lucifer or thin splint of wood into the reservoir of the 
lamp. A light blue flash of flame, visible within the aper- 
ture, and scarcely audible explosion will result if the mix- 
ture of air and vapour in the upper part of the reservoir is 
inflammable. Such an experiment is as harmless as that 
of lighting raisins soaked in brandy inthe game of Christ- 
mas snap dragons. If the explosive mixture is met with, 
the oil is dangerous, as the ignition of the mixture may 
occur when the aperture is closed by the brasswork; in that 
case the resulting flame and expanded products of combus- 
tion, having no other vent, will escape by bursting the 
reservoir, and scattering the oil, to the risk of a conflagra- 
tion, if the oil catches light. If there is no explosive mix- 
ture in the reservoir the oil is safe, and will even extinguish 
the test-flame if the latter is plunged beneath the surface. 
Experiments of this sort would, of course result in a demand 
which, passing on from mouth to mouth, would at last pro- 
duce the supply of safe oil from the refiner. But we must not 
expect such tests to be performed to a commensurate extent 
by the public. We must turn to the trade in this matter, and 
look to them to adopt this experiment, in its, to them, 
easier form of the test-tube and thermometer, already de- 
scribed, and to regard no refined oil as marketable if it 
