CONCERNING A STUDY OF KEROSENE OILS BY PHYSICAL 
METHODS. 
BY G. W. STEWART. 
Kerosene oil lias been the subject of investigation for a number of 
years, and with the result of improved processes in refining and an il- 
luminant which fulfills a definite need quite satisfactorily. But these 
investigations have been conducted almost wholly from the point of 
view of the chemist. This is really not surprising when the complex 
character of the oils and the necessarily very complex character of the 
chemical changes in the burning flame are considered. To secure any 
practical results by physical methods seems at first sight to be hopeless, 
for does not the illuminating power of the oils depend upon the con- 
stituents, and how is it possible do obtain any information save by 
means of a knowledge of these compounds? Yet the investigations 
concerning which this note is presented have given definite results, and 
assure us that a further investigation along physical lines is highly de- 
sirable. 
My object in this account is to give some of the results without 
a detailed description of the experiments. The work will be published 
with sufficient details in the near future. 
The first problem attacked was the ‘illuminating power” of the oils. 
It should be stated at the outset that the variations in illuminating 
power in the kerosene oils found on the market are not sufficient to 
make the subject of illuminating power the most important field of in- 
investi^ation. Such a study is, however, one of the most 'difficult. (The 
maximum variation in the oils studied is 17 per cent). One of the 
first precautions observed was to secure a method of comparing the 
illuminating power of the different samples under the same conditions. 
The conditions include the lamp, the wick, the size of the flame, the 
barometric pressure, the humidity, etc. The variation in the conditions 
of the atmosphere was eliminated by the use of a comparison sample. 
Whenever tests were made, observations were taken upon the sample 
used as a standard of comparison, and all results reduced- to the same 
basis. 
