4 The Bulletin. 



The iodine number should be low, as a high iodine number indi- 

 cates unsaturated hydrocarbons, which cause a yellowish flame. 



Photometric tests should show a steady flame of good quality and 

 good illuminating power. The illuminating power of the best oils 

 does not drop appreciably in a twelve-hour burning test. 



The various tests should he taken together in order to get a fair 

 estimate of the sample. 



METHODS OF ANALYSIS. 

 t 



The viscosity was determined with an Engler-Ubbelohde viscosime- 

 ter for illuminating oils. The viscosimeter for lubricating oils gave 

 viscosities from 1.07 to 1.13, except that the viscosity of sample ]Sro. 

 33 was 1.17. 



The distillation test was carried out by the continuous Engler 

 method as modified for the ofiicial residue test : 



"A weighed Engler flask, containing 100 cc. of illuminating oil, is protected 

 by a layer of asbestos on the bulb and neck. A thermometer is inserted in the 

 usual way so that the upper part of the mercury bulb is opposite the lower 

 part of the side neck of the flask. The distillation is conducted in the regular 

 manner over a Bunsen flame, the heat being regulated so as to distil the oil as 

 evenly as possible over a period of approximately 25 minutes. The time of 

 distillation is counted from the time the first drop leaves the condenser until 

 the burner is finally extinguished and removed. The temperature is brought 

 up to the actual corrected temperature of 570 degrees Fahr. a second time 

 after cooling at least 30 degrees Fahr. The corrected temperature is obtained 

 by applying the usual formula for correction for the exposed part of the ther- 

 mometer. The total period of heating shall not exceed 45 minutes. The 

 residue in the flask is reported in per cent by weight." 



The temperature was not corrected at 250° C. With the exception 

 of sample ^o. 33, the distillation began at 150° to 160° C. and the 

 distillate below 200° C. varied from 26 per cent to 45 per cent. 



Sulphur was determined gravimetrically after burning the oil. 



The degree of refining was determined by comparing the color 

 with Bismarck brown after shaking 100 cc. of the sample with 40 cc. 

 of 1.73 sulphuric acid for two minutes. (See Gill's Oil Analysis, 

 5th Edition, pp. 23-24, and Holde, 3d Edition, pp. 61-62.) Oils 

 giving a color above 8 are classed as "poorly refined." The test was 

 not applied to the red oils. 



The iodine number was determined on a one-gi'am sample by the 

 official Hanus method (Bulletin No. 107, rev.. Bureau of Chemistry, 

 pp. 136-137). The results agree approximately with the oils absorbed 

 by fuming (98.0 per cent) sulphuric acid according to the method 

 of Kramer and Bottcher (Holde, p. 36 and pp. 63-66). By the lat- 

 ter method from 6.0 per cent to 13.4 per cent of the oil was absorbed. 



The photometric test was made by measuring the candle power at 

 the end of the first and twelfth hours after lighting, using as a com- 

 parison light a standard electric bulb at 4 watts per candle power. 

 The lamp had a glass reservoir of 600 cc. capacity and was fitted 



