THE MODE OF TESTING MINERAL OILS USED FOR LAMPS. 647 
with air, may form an explosive mixture in the immediate neighbourhood of the flame. 
In other lamps the pin is fitted in such a way that there cannot be any accumulation of 
oil-vapour near the flame, and there is not any communication between the flame and 
the reservoir, except through the tube in which the wick is placed. The latter form of 
lamp is in this respect much preferable to the other, aDd it is the only kind of lamp in 
which a very volatile oil can be burnt. But the fact of most importance in regard to 
the determination of the inflammability of mineral by the “ flashing-point ” is that the 
result obtained with a given sample of oil in this way, may vary several degrees, accord¬ 
ing to the mode in which the test is applied. If the surface of the oil is freely exposed 
to the air, in an open vessel, such as a shallow capsule, the firing-point of the vapour 
may be found to be some degrees higher than when the test is applied in a small wide¬ 
mouthed bottle, or in the reservoir of a paraffin lamp, half-filled with the oil. 
The difference before referred to between the temperature at which, in testing oil to 
ascertain the firing-point, the temporary flash takes place, and that at which the oil itself 
takes fire and continues to burn, also depends very much upon the mode in which the 
experiment is made. With a shallow, open basin, the difference is much less than it is 
when a wide mouthed bottle or a small lamp-reservoir is used. The more freely the 
surface of the oil is exposed to the atmosphere, the lower will be the temperature at 
which it takes fire; while, on the contrary, the less freely it is exposed, the lower will 
be the temperature at which the vapour given off takes fire. _ Consequently, in fixing a 
point of temperature as the minimum at which either the oil itself or the vapour it gives 
off, should take fire by contact with flame, it is necessary that the conditions under which 
the experiment is to be made should be precisely defined, and that the testing of oils 
should always be conducted in the same manner, so as to obtain uniform and correspond¬ 
ing results. " It is, of course, desirable, also, that these conditions should assimilate as 
much as possible to those obtaining in a paraffin-oil lamp when it is burning, since any 
danger that might arise from the inflammability of the oil would exist chiefly in the 
ordinary use of this material in lamps. 
It is singular, however, that the Act passed in 1863, to amend the Petroleum Act of 
1862, prescribes a mode of testing mineral oil under conditions that are the direct oppo¬ 
site of those prevailing in the use of the oil. While a mineral oil lamp is in use, the oil 
is heated in a elosed vessel partly filled with air, which thus becomes charged with oil- 
vapour, proportionately to the volatility of the oil, or part of it, and to the temperature 
the oil is raised to. Any oil-vapour given off is confined; and, if the oil be sufficiently 
volatile or sufficiently heated, such formation of oil-vapour may result in the production 
of an explosive mixture within the oil-reservoir of a lamp. 
On the contrary, in testing mineral oil according to the directions given in the Schedule 
appended to the Act of 1868, the oil is heated in an open vessel, with the surface freely 
exposed to the atmosphere. Under these conditions, any oil-vapour formed is liable to 
diffuse away into the surrounding air, and become thus so much diluted as to lose its in¬ 
flammable character. Consequently, the temperature at which the momentary flash 
takes place with any given sample of oil when a flame is brought near its surface, will 
be, for that reason, higher than when the escape and dilution of the oil-vapour is pre¬ 
vented—as in a paraffin lamp, or by making the test in a partially closed vessel. 
A still more serious interference with the results of the test would be experienced if 
the operation were conducted in a place exposed to draughts; and probably much of the 
discrepancy in the results of oil-tests has arisen in this way. 
But there is yet another source of discrepancy in the results of tests made in accord¬ 
ance with the directions of the Petroleum Act, and one which is of far more importance, 
because no notice is taken of it in those directions, and, consequently, its influence may 
be greater or less according to accident or to the practice of the operator. This source 
of error lies in the rate at which the oil is heated, or the time occupied in making a 
test. All that the Act directs in this respect is that “ a small flame shall be applied to 
the bottom of the outer vessel,” which serves as a water-bath, and that this vessel “shall 
be filled with cold, or nearly cold water.” Both these directions are extremely vague. 
But, leaving out of consideration any difference that might arise from using “ cold ” 
water at 40° F., or “ nearly cold ” water at 70 3 F., the rate of heating the oil from either 
temperature up to 100° F. has such an influence on the result obtained that there may 
be a difference of quite 5° in the flashing-point of a given sample of oil, accordingly as 
the heating is made to occupy a longer or shorter time; and this will be the result, even 
