ach2 On the Igniting Point of Petroleum. 
flammable vapour, some amount of stirring or agitation 
should be practised. 
The distance of the test-fame from the petroleum may 
cause variation in observing the igniting-point. The 
amount of vapour evolved in a given time may be sufficient 
to form a small quantity of explosive mixture near the sur- 
face of the petroleum, but not at an inch or two from the 
surface; hence an operator dipping the test-flame down to 
the surtace of the liquid, or even below the surface, to see 
if the oil will extinguish instead of being ignited by his 
flame, will give a lower igniting-point of the vapour than 
an operator who only brings his test-flame within an inch 
of the surface of the oil in a basin, or only just within the 
mouth of a bottle or gallipot. In an experiment with the 
“Ladoga” oil, this cause of error made a difference of 4° 
in the observed igniting-point. 
The amount of petroleum operated on in comparison with 
the size of the vessel in which the operation is performed 
will, of course, influence the observations, for the reason just 
stated. An operator making his experiments in a bottle 
or deep beaker or gallipot, and always introducing the 
test-flame only just within the mouth, will sooner meet with 
the explosive mixture if the vessel is three-fourths full, than 
if that mixture has first to fill the upper part of a vessel 
only one-fourth occupied by the liquid. 
These, then, are causes, amply sufficient in number and 
nature, of the variation in the igniting-point of petroleum, 
as observed by different experimenters, or by the same ex- 
perimenter at different times. They fully explain the fact 
that I have found a specimen of petroleum to ignite at 
almost any degree between 78° F. and 124° F. Before pro- 
ceeding to suggest a modified method of observation which 
shall give constant results, it may be instructive to give a 
table of the igniting-points of several specimens of refined 
petroleum recently circulating in wholesale and retail com- 
merce’™* the experiment on each specimen being taken un- 
der three conditions :—first, heating in an open earthenware 
bowl, of the form, &c.. already described; second, heating 
in a two—ounce, wide-mouthed, white glass phial, without 
shaking the petroleum; and third, as the second, except 
that before introducing the test-flame, the stopper was in- 
serted in the bottle, and the whole well shaken for three or 
* For the former I have to thank Messrs. Rose, Graham, and 
Wilson, of Threadneedle street. 
