324 TRANSACTIONS OF SECTION A. 



3. On the Methods and Apparatus used in Petroleum Testing. 

 By J. A. Harker, D.Sc, F.R.S., and W. P. Higgins, B.Sc, A.R.C.S. 



An International Commission has recently been appointed to discuss the 

 unification of the methods and apparatus used in petroleum testing. The 

 National Physical Laboratory has joined in the work proposed by undertaking 

 an investigation into the general methods and apparatus for the determination 

 of the flash-point of petroleum used for illuminating purposes. The apparatus 

 first studied include the well-known Abel tester, the official type used in this 

 country, the Abel-Pensky as used in the British Colonies and India, and the 

 German form of the same instrument. It has long been known that for some 

 hitherto unexplained reason the Abel-Pensky apparatus gives a higher result 

 for the same oil than the Abel ; therefore in the testing of a cargo of oil it 

 becomes a matter of importance as to which form of apparatus is used for the 

 purpose. The object of the present paper is to determine the cause of these 

 considerable differences before any steps are taken as to international agreement 

 upon the matter. The Abel apparatus consists of a cylindrical brass oil-cup 

 supported in an air-jacket surrounded by a water-bath adjusted to a definite 

 temperature. The oil-cup is fitted with a cover which carries a thermometer 

 divided into whole degrees Fahrenheit. The oil whose flash-point is required 

 is filled into the oil-cup to the level indicated by a projecting point. Withdrawal 

 of a sliding cover depresses a small test-flame a specified distance into the 

 space above the oil. An ivory bead of specified diameter 6erves as a gauge in 

 adjusting the flame to a uniform size. In making a determination of the flash- 

 point the test-flame is applied to the vapour by pulling the slide as the mercury 

 crosses each degree mark and this operation is timed by a pendulum or metro- 

 nome. The exact timing was deemed to be of considerable importance when 

 the original specification was issued. When the apparatus was adopted in 1880 

 by the German Government as their official standard, alterations in some details 

 and dimensions were introduced by Pensky. The chief modifications consisted 

 in the addition of a clockwork mechanism to open and close the slide auto- 

 matically in the specified time and in the substitution of the Centigrade for the 

 Fahrenheit temperature scale employed in the Abel apparatus. In this form 

 the apparatus is known as the ' Abel-Pensky ' type. The British Colonies 

 and India in their official type (which is not legally recognised in England) 

 adopted Pensky's automatic mechanism for the cover, but retained the dimen- 

 sions and Fahrenheit scale of the English specification. Some preliminary in- 

 vestigations were made in order to study the effect of definite variations in the 

 experimental conditions. The results obtained showed that large differences of 

 temperature existed throughout the oil-cup and vapour-space above it at any 

 stage in an experiment. To study adequately the distribution of temperature 

 throughout the apparatus electrical means were used and the differences at 

 various points were determined by means of sensitive, minute thermocouples 

 of appropriate type. The general theory of flash-point determination depends on 

 the hypothesis that flashing takes place when the space above the oil contains 

 a definite percentage of oil- vapour mixed with air. This condition will be 

 reached for a definite temperature of the oil from which evaporation is taking 

 place, and it is generally assumed that it is this temperature which is given by 

 the thermometer as the flash-point. The rate of evaporation, however, depends 

 on the temperature of the surface of the oil, and the present observations show 

 that this varies from point to point and at any moment differs appreciably from 

 the thermometer reading. The temperature distribution depends in part on the 

 exact form of the apparatus and on the relative amounts of heat reaching it 

 from different sources. The difference above referred to frequently amounts to 

 5° F. and over, and is not the same, at corresponding readings of the thermo- 

 meter, in the different types of apparatus. Measurements of the heating effect 

 of the test-flame were also made and these showed that the size of the test-flame 

 was of considerable importance— a fact which hitherto does not seem to have 

 been adequately recognised. This heating effect is manifested mostly in the 

 upper parts of the oil-cup, tending to raise the temperature of these parts above 

 that of the rest of the oil. Now in the Abel-Pensky apparatus the cover con- 

 tains a much greater amount of metal than in the Abel, and hence the former 

 requires a greater amount of heat to raise its temperature by any definite 



