Scientific Lectures. 205 



but should be -protected from currents of air, when in use, by a 

 screen. 



The closed tester, or " pyrometer," I consider very unreliable, at 

 least for determining the burning point, as the mass of metal (brass) 

 over the oil is very liable to become heated by the burning vapor 

 after the flashing point has been reached. The result of this is the 

 igniting of the overheated petroleum at the surface long before the 

 thermometer records is actual temperature. I have seen the oil 

 ignite while the thermometer stood 20° below its proper burning point. 



When heat has been applied to the water until the thermometer 

 . has risen to about 90° Fahrenheit, a very small flame shall be quickly 

 passed across the surface of the oil on a level with the wire. If no 

 pale blue flicker or flash is produced, the application of. the flame is 

 to be repeated for every rise of two or three degrees in the thermome- 

 ter. When the flashing point has been noted, the test shall be 

 repeated with a fresh sample of the oil, using cold, or nearly cold 

 water as before ; withdrawing the source of heat from the outer ves- 

 sel when the temperature approaches that noted in the first experi- 

 ment, and applying the flame test at every rise of two degrees in the 

 thermometer. 



The mere fact that an oil does not take fire from a match, does not 

 prove that it is a safe oil, if it is at the temperature of the atmos- 

 phere. The oil refiners have selected the temperature of 100° 

 Fahr. as the standard temperature at which a good oil should not 

 emit an inflammable vapor, and 110° as the temperature at which 

 it should not take fire. If it will not take fire at 65° it may not 

 be safe ; for at 70° we may reach its burning point. This test will, 

 however, readily distinguish between a good kerosene and the gaso- 

 line or falsely called " safety oils " which burn at all temperatures. 



Spontaneous Combustion. 



Mechanical division has a great influence upon combustion. If we 

 reduce a substance to a fine powder, we frequently find that it will 

 combine with oxygen at a much lower temperature than otherwise. 

 I have here tubes containing metallic iron in so finely divided a con- 

 dition that it takes fire spontaneously in the atmosphere, as you see. 



Charcoal reduced to fine powder is liable to spontaneous combus- 

 tion. In gunpowder factories it is necessary to be very careful not 

 to expose powdered charcoal to the atmosphere. Porous substances 

 have the power of condensing gases, which raises their temperature ; 

 and finally it rises to such a point that combustion takes place. Tow, 



