76 BULLETIN 1216, U. S. DEPARTMENT OF AGRICULTURE. 
The accuracy of the thermometer when delivered to the pur- 
chaser shall he such that when tested at full immersion the 
maximum error shall not exceed the following: 
Prom 0° to 200° G. : 0.5° C. 
From 200° to 300° C. : 1.0° C. 
300° to 375° C. : 1.5° C. 
The sensitiveness of the thermometer shall be such that when 
cooled to a temperature of 74° C. below the boiling point of water 
at the barometric pressure at the time of test and plunged into 
free flow of steam the meniscus shall pass the point 10° C. below 
the boiling point of water in not more Than six seconds. 
10. The retort shall be supported on a tripod or rings over two 
sheets of 20-mesh gauze, 6 inches square, as shown in Figure 12. 
It shall be connected to the condenser tube by a tight cork joint. 
The thermometer shall be inserted through a cork in the tubula- 
ture with the bottom of the bulb one-half inch from the surface 
of the oil in the retort. The exact location of the thermometer 
bulb shall be determined by placing a vertical rule graduated in 
divisions not exceeding one-sixteenth inch back of the retort when 
the latter is in position for the test, and sighting the level of the 
liquid and the point for the bottom of the thermometer bulb. The 
distance from the bulb of the thermometer to the outlet end of 
the condenser tube shall be not more than 24 nor less than 20 
inches. The burner should be protected from drafts by a suitable 
shield or chimney (see fig. 12). 
Method of test 11. Exactly 100 grams of oil shall be weighed into the retort, the 
apparatus assembled, and heat applied. The distillation shall be 
conducted at the rate of at least 1 drop and not more than 2 
drops per second, and the distillation collected in weighed re- 
ceivers. The condenser tube shall be warmed whenever necessary 
to prevent accumulation of solid distillates. Fractions shall be col- 
lected at the following points: 210°. 23."°. 270°, 31.")°, and 355° C. 
The receivers shall he changed as the mercury passes the dividing 
temperature for each fraction. When the temperature reaches 
355°, the flame shall be removed from the retort, and any oil which 
has condensed in the offtake shall be drained in the 355° fraction. 
The residue shall remain n the retort with the cork and the 
thermometer in position until no vapors are visible; it shall then 
be weighed. If rhe residue is to be further tested, it shall then 
be poured directly into the brass collar used in the float test" 
or into a tin box and covered and allowed to cool to air tempera- 
ture. If the residue becomes so cool that it can not be poured 
readily from the retort, it shall be reheated by holding the bulb 
of the retort in hot water or steam, and not by the application 
of flame. 
For weighing the receivers and fractions, a balance accurate to 
at least 0.05 gram shall be used. 
During the progress of the distillation the thermometer shall 
remain in its original position. No correction shall be made for 
the emergent stem of the thermometer. 
When any measurable amount of water is present in the dis- 
tillate it shall be separated as nearly as possible and reported 
separately, all results being calculated on a basis of dry oil. 
When more than 2 per cent of water is present, water-free oil 
shall be obtained by separately distilling a larger quantity of oil. 
returning to the oil any oil carried over with the water, and using 
dried oil for the final distillation. (See "Water test") 
Specific Gravity of Fk actions. 
12. As specific gravity is an absolute physical determination, 
any recognized method which can be applied to the quantity and 
quality of material at hand to be tested must be considered satis- 
factory. The following methods are convenient and accurate 
means for the relatively small amounts of oil available in deter- 
mining specific gravity of fractions to be tested. 
16 See " Float test of residue,'" p. 78. 
