as the presence of a small quantity of highly volatile hydrocarbon, presumably too small in quantity to sensibly affect the flash-point, increases the pressure in the apparatus. The conclusion is, however, expressed that oils whose vapour pressure at 15° C. is not greater

Test-tube method of testing.

than is represented by a column of 64 millimetres of

water may be considered safe.

When only a small quantity of oil is available for testing, the flash-point may be approximately determined by slowly heating the oil in a test tube 4 or 5 inches in length by about three-quarters of an inch in diameter, and inserting the burning end of a piece of thin twine into the mouth of the tube at intervals

after agitating the oil. The tube should be about one-third filled with the oil, and a delicate thermometer used to take the temperature. In this way results which do not differ materially from those furnished by the Abel instrument may be obtained after a little practice.

of tanks, &c.

flame.

tal investiga

tion.

The subject of the detection and measurement of Testing petroleum vapour in the atmosphere of tanks and atmosphere other spaces engaged the attention of one of the authors about eight years ago. He was at one time accustomed to employ an alcohol flame in testing for Use of petroleum vapour, but becoming impressed with the alcohol importance of being provided with a delicate and thoroughly trustworthy means of determining the proportion of inflammable vapour present, he was led to investigate the subject. After experimenting with Experimenvarious methods of testing, with results given in a paper published in the Minutes of Proceedings of the Institution of Civil Engineers,* he devised, in consultation with Professor Clowes and Mr. Robert Redwood and with the assistance of Messrs. W. J. Fraser & Co., a testing apparatus in which the hydrogen flame is employed. It is well known that a non-luminous flame burning in a space from which light is excluded, in air containing a small proportion of inflammable gas or vapour, is seen to be surrounded by a faint halo, which is termed the "flame-cap." The advantage "Flamepossessed by a hydrogen flame over other flames in cap." point of sensitiveness when thus employed as a test of hydrogen for inflammable gases has long been recognised; and flame. Mr. Pieler,‡ in 1883, after referring to the previous

* "Redwood on the Transport of Petroleum," Min. of Proc. Inst.

C.E., exvi. (1893-94), Part II.

+ English Patent No. 187, 1893.

Ueber einfache Methoden zur Untersuchung der Grubenwetter. Aachen, 1883.

Advantage

recommendation of the hydrogen flame for testing by Messrs. Mallard and Le Chatelier,* described an apparatus in which the gas generated in a Dobereiner lamp was burned in a specially constructed test-lamp, for use in testing samples of air from the workings in coalmines. Professor Clowes is entitled to the credit of having devised a form of miners' testing-lamp, which depends for its success upon the employment of hydroRedwood's gen stored under pressure. In the Redwood testinglamp compressed hydrogen is also used.

Clowes' miners' test-lamp.

test-lamp.

Description.

Fig. 11.

The complete appliances are shown in Figs. 11 and 12. They consist of the lamp A, the reservoir of compressed hydrogen B, and the sampling vessel C, in which the sample of air for examination is collected. The lamp is shown in section in Fig. 12. A is the hydrogen inlet tube with the regulating valve B, and is the hydrogen jet. D is the inlet tube for the

* Ann. d. Mines, 7th Ser., xix.

The apparatus is supplied by the sole makers. W. J. Fraser and Co., 98 Commercial Road East, London, E.

sample of atmosphere to be tested. The bore of this tube is greatly contracted, and immediately above the point at which the tube enters the base of the lamp is an arrangement of baffles, surmounted by three discs of wire gauze of at least 28 wires per lineal inch, or not

less than 784 openings per square inch, the flow of the gaseous mixture to the flame being thus regulated, and passage of flame into the collecting-vessel being prevented. The chimney E fits air-tight at the base, but is capable of vertical movement on an inner tube, the front of which is removed. The chimney is partly of metal and partly of glass, the metallic portion being

blackened inside, and on the glass window, lines corresponding with various heights of flame-caps may be marked. The top of the hydrogen jet-tube is 10 mm. (0.4 inch) below the bottom of the window. Attached to the base of the lamp is a telescopic support for a cloth, which envelops the head of the observer and excludes light when the testing apparatus is used in an undarkened room. The construction of the

collecting-vessel is shown in section in Fig. 12. A is the compression-pump, which is furnished with a metallic spring-piston, fitting the pump cylinder without the use of leather or other material, and lubricated. with plumbago. Surrounding the pump is an annular space, in which the sample of atmosphere is stored. B is a collar to which may be attached a flexible suctiontube of any desired length. C is a cock, to which is connected a copper tube conveying the sample to the test-lamp. The bore of this cock is very much reduced. D is a pressure gauge, and E a spring valve lifting at 30 lbs. pressure. FF are hinged brackets, on which the feet of the operator are placed while the pump is being worked. A handle is provided by which the cylinder can be conveniently carried. The capacity of the pump is 14.84 cubic inches, and of the annular space 169 14 cubic inches, thirty double strokes of the pump being required to charge the vessel to a pressure of 30 lbs. per square inch, when it will contain foot of the atmosphere sampled. It is desirable that the collecting-vessel should be fitted with a relief valve, as the apparatus is often used in places in which the dial of the pressure gauge cannot be easily seen. hydrogen cylinder may be of any desired size, but what is known as a 5-feet cylinder is of convenient dimensions. When charged to the usual pressure of 120 atmospheres, it holds enough gas to supply a 10-mm. flame for ten hours, and is quite portable.

cubic

The