502 WRITINGS OP JOSEPH HENRY. [1875 



to render this change desirable in an economical point of 

 view. In the meantime experiments had also been made 

 in France and England for the purpose of introducing min- 

 eral oil as a light-house illuminant, but it was not until 

 1873 or 1874 that the result was entirely satisfactory. 



The process of manufacturing the oil has been very much 

 improved in this country of late years, and there are now 

 several companies which profess to produce oil entirely safe, 

 and otherwise suitable for light-house purposes. In view of 

 further experiments with mineral oil, an advertisement was 

 inserted in the papers in 1874 requesting manufacturers to 

 send samples of their oils to be tested at the Light-House 

 depot at Staten Island, and in accordance with this nume- 

 rous specimens were received and submitted to examina- 

 tion. 



The first test to which the oils thus furnished were sub- 

 jected was that of flashing ; that is, the determination of the 

 temperature at which the oil gives off a vapor which will 

 flash into a flame on the approach of a small taper, or in 

 other words which indicates the rise of a vapor which mixed 

 with atmospheric air will tend to produce an explosion. 

 The flashing temperature differs however from that at which 

 the liquid takes fire as a whole. This will be understood if 

 we suppose that two liquids have been mixed together, a 

 light and a heavy one ; the flash in this case will be due to 

 the vapor from the lighter mixture, while the burning is 

 due to the temperature at which the compound is fired. To 

 make this flashing test requires considerable precaution. 

 The oil to be tried is gradually heated by a spirit-lamp 

 in a water-bath, a sensitive thermometer being suspended 

 in the oil with the bulb slightly below the surface; the heat 

 of the water is very slowly increased by moving from time 

 to time the spirit-lamp from under the basin of the water- 

 bath which contains the oil, and the point of flashing is 

 obtained by passing over the surface of the oil a small flame 

 until the first indication of flash is observed. The flame 

 should not be so large as to heat the surface, and is best pro- 



