320 On the Cause of Explosions in Lamps. 
sooner or later, a gaseous mixture which, should it catch 
fire, will burst the lamp and scatter the oil, to the possible 
injury of property and danger to life. Practically, I have 
only obtained lamp-explosions with oils which give off 
vapour below 100 degrees, when examined in the test-tube 
in the manner above described. Unfortunately, fourteen 
out of every fifteen specimens of “crystal oil,” “ photogen,” 
“ American paraffin oil,’ and other varieties of petroleum 
now sold for illuminating purposes, generate the explosive 
mixture at temperatures much below 100 degrees, and 
hence are dangerous. The same remark does not apply to 
Young’s paraffin oil, but there is reason to fear that this 
mineral oil is often diluted with dangerous American oils. 
Other home-made mineral oils also vary in explosibility. 
The explanation of the occurrence of dangerous mineral 
oil in trade was fully traced out in my last paper; three 
sentences, therefore, will be sufficient on this head. Crude 
petroleum having been found to be inflammable at common 
temperatures, an Act was passed in 1862, forbidding, except 
by licence, the storing of petroleum and its products near a 
dwelling-house or warehouse in larger quantities than forty 
gallons, unless proof was forthcoming that it gave off no 
“inflammable vapour at a temperature of less than one hun- 
degrees of Fahrenheit’s thermometer.” Unfortunately, this 
was taken to mean, and probably did mean, the petroleum 
itself should not ignite under 100 degrees when heated by 
a flame or a warer-bath in an open saucer or a small bowl, 
—a most fallacious test, first, because the time employed 
in the operation, the amount of stirring practised, the form 
of the vessel, its arrangement over the source of heat, the 
quantity of liquid used, and the distance of the test-flame 
from the surface of the petroleum, all influenced the result 
to the extent of causing two observations to vary from 4 to 
40 degrees from each other; and, secondly, because the 
experiment thus performed invariably gave the igniting- 
point many degrees higher than the temperature at which 
_inflammable vapour was evolved,—than the temperature, 
therefore, at which the oil was dangerous. Shippers, mer- 
chants, brokers, adopted this fallacious mode of testing; 
oil changed ownership under a warrant that it would not 
ignite under 100 degrees, not, however, without occasional 
litigation caused by the uncertainty of the test, and hence 
the market was and is supplied with oil which emits in- 
flammable vapour in most cases below go and often below 
80 degrees, } 
