904 TEAi\SACTIO?rS OF THE AMERICAN INSTITUTE. 



The oils were lieated in a small capsule over a water or paraffine 

 bath, a thermometer being inserted in the oil, and a thin burning 

 wick being held over the same. 



The petroleum oils which were experimented upon, were very 

 differently obtained, a part of them were gathered directly from the 

 cooling worms in relineries, others were obtained by fractional dis- 

 tillation in small retorts, and still others by evaporatio n of speeili- 

 cally light mixtures. 



The two first samples of the density of 0.643 and 0.686 already 

 took fire at 5.8 degrees Fahrenheit, henceforth the inflammability 

 diminishes till the density of 0.822 is reached. From this point we 

 again see it increase. This remarkable fact is easily explained, when 

 we consider that the high temperature which is necessary to distil 

 the oils of 0.822, is suflicient to produce a partial decomposition of 

 the higher boiling oils in the retort. 



This admission is sufliciently confirmed by the experiment. "When 

 the distilled oils had reached the density of 0.822, the fan under the 

 retort was drawn out. In producing a light oil of 0.800, distillers 

 generally gather only the portions which come up to this point ; the 

 first fractions which are used with the illuminating oil possess a 

 specific weight of 0.750 ; the mixture does not then take fire below 

 96.7 degrees Fahrenheit. The remainder in the retort may be 

 heated to 343.4 degrees Fahrenheit before it is inflamed by a burning 

 wick. When, however, after the distillate had reached the specific 

 gravity of 0.822, the heat was increased, as it is done for the produc- 

 tion of lubricating oils ; the inflammability was also increased, as is 

 seen from the foregoing table. 



Refijied parafiine of a melting point of 129.2 degrees Fahrenheit 

 could be heated to 429.8 degrees Fahrenheit ; it then took fire, but 

 without a prior decomposition being noticed, which obviously had 

 taken place in the distillation of the heavy oils and crude oil contain- 

 ing paraffine masses. 



The schist oil samples were obtained from a distillation on a small 

 scale. The same was carried out in a cast-iron retort of two and one-half 

 gallons capacity on naked fire. The oils were purified and from Autun. 

 It is striking that the latter arc a great deal more inflammable than the 

 petroleum oils of the same density, Pruf. Maroc, of Stuttgardt, also 

 indicates the inflammability of a schist oil, which he does not desig- 

 nate further than as being at 63.5 degrees Fahrenheit, 



It is highly probable that a similar decomposition goes on in the 



