the temperature of the oil in an ordinary glass oil-lamp ranges from 76° to 98° F., and in a metal lamp from 76° to 129° F., the lower limits being for rooms heated between 73° and 74° F., and the higher for a temperature of 90° to 92°. It is, therefore, evident that an oil giving off explosive gases at less than 100° F. must be dangerous, and even at 110° F. an accident might occur, but only in exceptional circumstances.
The oils must, therefore, stand a certain test, called the "flashing test," which consists in heating them, preferably, in a thin metal or glass cup which holds the oil, and is itself placed in another vessel full of cold water, which is gradually heated by a small spirit-lamp. The bulb of a thermometer is kept well immersed beneath the surface of the oil, draughts are to be avoided, and the heat very slowly raised. From time to time, as the flashing-point is approached, the temperature is noted, and a very small flame, as a gas-jet issuing from a glass tube drawn to a fine point, is quickly passed across its surface, taking care not to touch the oil. A faint blue flame will flash across the oil when it reaches a temperature at which explosive gases are given off. Although it is generally agreed that the temperature should be very gradually raised, fifteen minutes being allowed for a test, yet Calvert (Chemical News, May, 1870) states that an oil which flashed at 90° F., after fifteen minutes, showed a flashing-point of 101°, when thirty minutes were consumed in making the test. Oil of 100° is not safe absolutely. There is another test called the burning-test, the point at which an oil will take fire and burn; it is from 10° to 50° F. above the flashing-test (Chandler), and is of little value in determining the safety of an oil, because, as already shown, the addition of one per cent, of naphtha will lower the flashing-test 10° in a good oil, while it would not materially affect the burning-point. From the directions already given for testing oil any one can readily make the test, and in view of the large number of unsafe oils sold it is very important that such tests should be made before using an oil not known to be safe.
The subject of refining petroleum may be dismissed with a few words more about "cracking" oils. It is the object of the refiner to make as much illuminating oil as possible, and to do this advantage is taken of the fact that, when the vapors of heavy oils are heated above their boiling-points, carbon is deposited, and the condensed hydrocarbons resulting have a less specific gravity. This decomposition is technically called "cracking," and it was observed long ago that in distilling the heavier oils lighter hydrocarbons were obtained during the first stages of the operation, even when not wanted. Cracking can be accomplished by distilling the oils under pressure, or, as is the case in the very large stills now employed, by allowing the vapors of the heavier hydrocarbons, on condensing, to flow down again upon the now hotter oil in the still, whereby they are cracked, depositing carbon. By carefully adapting the heat to the changing