ANALYSES OF THE OILS. 
207 
onds) the results must be divided by 25. Through the instrument which I used water 
ran in 21b seconds. Therefore I give two sets of results; one set consists of the number 
of seconds it took the oil to run through the instrument which I used; the other set, 
the same number of seconds divided by 21b. The data of the last set approximately 
tell how many more times viscous the oil is than water at 15° C. (about 60° F.). Just 
before the test is made the level of the oil is made even with the tip. 
Flash point .—Nothing can be more inaccurate than the open tester, since it may be 
manipulated at will to give results varying by 10°. The most accurate of the open 
testers is liable to great variations in its results from drafts of air. The Abel closed 
tester gives results which may be duplicated within 1° or 2°. It is the English stand¬ 
ard, and with the Pensky attachment it is the German standard. It seems to be per¬ 
fect in every respect for its purpose, i. e., for determining the flash from 60° F. to 
160° F. (Why not adopt it as the standard for California?) If the instructions for open¬ 
ing the slide are carefully followed, the Pensky clockwork for opening the slide is not 
necessary; it was invented to do away entirely with any tampering with results. In 
making the above determinations of flash point the directions given in Redwood’s 
Manual were followed exactly. Since the flash was taken with reference to safety in 
transportation, the test was not made on those samples flashing below 15° C. (60° F.) 
and above 70° C. (168° F.). 
Calorific value .—All the determinations were made with thte Atwater bomb. The 
thermometer used was a Beckman, graduated in one one-hundredfhs, whose accuracy 
I tested. The basis was the average between Berthelot’s calorific values for cane 
sugar and naphthalene. I carefully determined the value of these two substances. 
The naphthalene used was made by De Haen, Germany; its boiling point was between 
79° and 80°, and it was melted before using. The sugar was three times crystallized 
cane sugar, whose composition by organic combustion agreed with the theoretical by 
0.03 on the II and 0.05 on the C; in other words, perfectly pure sugar. Before using, 
it was dried at 50° in the oven. My average of several results on sugar and naphtha¬ 
lene had to be increased by 1.08 per cent to make it the same as Berthelot’s average for 
sugar and naphthalene, and all my results on petroleum are increased by the factor 
1.0108, which is evidently the correction factor of the bomb calorimeter plus my per¬ 
sonal equation. My results are certainly within 0.5 per cent of the absolute, and on 
repetition the duplicates agree within 0.3 per cent. 
Samples Nos. 1, 2, 3, 4, 6, 7, 9,10, 14,17,19, 25, 29, 38, 40, 41, 42, 43, 46, and 48 were 
taken from the top of the cans after the oil had stood for three or four months, and the 
calorific value found without preliminary treatment. Since Nos. 3, 4, 17, 29, and 40 
had water present, the calorific value obtained for these was increased by the same 
number of per cent as the oil held water; in other words, the calorific value was cal¬ 
culated on the dry sample. All other calorific values that were made of samples con¬ 
taining water first had the water removed by distillation. The calorific value of an 
oil containing water is useless for most purposes unless the percentage of water is given 
also; for how can anyone get concordant results on a sample half water? (See sample 
No. 30.) Oil is bought by the gallon and not by the pound, and the column headed 
“Calorific value per c. c.” gives the comparative value of the oils by volume. 
Sulphur .—The sulphur was determined from the sulphuric acid left over from the 
combustion in the platinum cup. The cup or lining as well as the lid of the bomb was 
carefully washed into a beaker, the liquid filtered to about 100 c. c., BaCl 2 added, and 
the precipitated BaS0 4 weighed and the sulphur calculated. The gases of combustion 
I found to contain no sulphuric acid. (And of course no S0 2 , since HN0 3 was present.) 
I passed the gases slowly through glass wool soaked in t n q KOH, and found almost no 
consumption of the alkalinity of the KOH, showing that all the H 2 S0 4 and HN0 3 
stuck to the lining. 
Distillation .—An ordinary ^-liter glass distilling flask was used. The exit tube is 
bent upward for 2 finches just after leaving the neck of the flask; this is to facilitate 
