10 BULLETIN 1308, U. S. DEPARTMENT OF AGRICULTURE 
chains the phenols should tend to resemble the fatty acids, the sodium 
salts of which possess marked power to dissolve and emulsify hydro- 
carbons. 
A second possible error may lie in the assumption that admixture 
of kerosene and cresol is unattended by any change in volume. 
EXPERIMENT 3(a) 
Cresol L B was distilled and again reboiled in an open flask to 
insure elimination of water. Its specific gravity at 25° C. was 
1.0330. It was then mixed with purified and dried kerosene of 
specific gravity 0.8063 in the proportion of 26.32 grams (=25.58 
cubic centimeters') cresol to 60.87 grams (=75.79 cubic centimeters) 
kerosene. On a purely additive basis the 87.19 grams of the mixture 
should have possessed a total volume of 101.37 cubic centimeters. Its 
specific gravity was found to be 0.8625 at 25° C. ; therefore its actual 
volume was 101.50 cubic centimeters, so that an expansion in volume 
on mixing at constant temperature occurred amounting to 0.13 cubic 
centimeter. 
EXPERIMENT 3(b) 
Cresol H B similarly treated showed an initial specific gravity of 
1.0231. Mixed with kerosene of specific gravity 0.8088 in the pro- 
portion of 25.22 grams cresol (=24.75 cubic centimeters) to 56.49 
grams kerosene (=70.12 cubic centimeters) the specific gravity of 
the mixture was 0.8638. The actual volume of the mixture at 25° C. 
was therefore 94.97 cubic centimeters, while the purely additive 
volume would be 94.87 cubic centimeters, or an expansion of 0.10 
cubic centimeter was effected by mixing. 
It appears, therefore, that the expansion in volume consequent 
upon the admixture of kerosene and cresol will cause the apparent 
volume of cresol to be between 0.10 and 0.13 cubic centimeter too 
great when about 25 cubic centimeters of cresol are present. Paraffin 
hydrocarbons are better solvents for higher boiling phenols than for 
the lower homologues ; consequently the error should diminish as the 
boiling points of the phenols ascend. 
A third, and probably the gravest and most erratic source of error 
in the Weiss-Hill method, lies in the incomplete elimination of water. 
Hill prescribes that — 
Where the sample to be tested contains water it should be removed by 
catching the first few cubic centimeters of distillate in a small separatory 
funnel containing a little saturated salt solution, separating the water and 
adding the oil to the rest of the distillate in the tar-acid funnel. 
In the writer's experience traces of water are nearly always pres- 
ent in the first portions of distillate, and further traces are likely 
to arise from decomposition toward the end of the distillation. 
Sodium-chloride solution is not an effective dehydrating agent. In 
fact, solid salt will separate from a saturated solution of sodium 
chloride that is shaken with an anhydrous kerosene solution of 
cresol. Therefore this method for the removal of water merely 
means that a certain equilibrium proportion is left in the sample, 
and the procedure may actually introduce a higher proportion of 
water into a sample than was originally present, which water, of 
course, will be finally reckoned as " tar acid." The writer's remedy 
