SAMPLING AND TESTING HIGHWAY MATERIALS. 
71 
made after twirling the funnel and allowing th< s to settle 
for a few minutes. The percentage shall be computed i».\ volume. 
The naphtha or light oil used shall be tested to determine free- 
dom from water. 
3. The accuracy obtainable by this method la within 0.1 per 
cent. 
4. Every precaution must be taken to insure a representative 
sample. Bituminous materials containing separated water are 
especially difficult to sample. 
METHODS OF SAMPLING AND ANALYZING CREOSOTE OIL. 
i A. S. T. M. 
standard methods, serial designation D 38-18 
modified.) 
slightly 
SAMPLING. 
1. Wherever the oil is being loaded or discharged by means of Sampling. 
a pump the following: method shall he used : 
A one-half inch sampling pipe shall be inserted in the line 
through which the oil is l=eimr pumped, on the discharge side of 
JUS| 
F^T 
I 
!B 
[III 
X [ 
ip^ , 
9.5 cm. 
Pig. 8. -Copper still. 
the pump, preferably in a rising ^e<•t!on of the pipe line. Tins 
sampling- pipe shall extend one-half way to the center of the main 
pipe and with the inner open end of the sampling pipe turned at 
an angle of 90° and facing the flow of the liquid. This pipe shall 
he provided with a pirn: cock and shall discharge into a receiver of 
50 to 100 gallons capacity. The plug cock shall be s,. adjusted 
that, with a steady continuous flow of the oil. the receiver shall 
be filled iu the time required to pump the entire shipment The 
receiver shall he provided with ;i steam coil sufficient to keep the 
contents at a temperature not exceeding 120° V. Immediately 
upon completion of the pumping, the content- of the receiver shall 
be very thoroughly agitated and a duplicate 1-quart sample taken 
immediately for the test. The amount of the drip sample col- 
lected shall he not le S < than 1 gallon for each 1,000 gallons of oil 
handled, except in the case of large boal shipments, where a 
maximum of luu gallons is sufficient. 
