CONTROL OF SAPONIFIED SOLUTIONS 15 
centimeter, which indicated that water was thus completely elimi- 
nated. After measurement the acid-extracted kerosene solution of 
phenols was extracted twice with a 10 per cent solution of sodium 
hydroxide, using 80 cubic centimeters and 60 cubic centimeters and 
allowing draining periods of two hours and one hour, respectively. 
The second extraction removed not more than 0.15 cubic centi- 
meter, which indicated that a third extraction would be superfluous. 
The decrease. in volume of kerosene was divided by 0.2515 to obtain 
the apparent volume percentage of phenols in the sample. Results 
obtained thus were as follows: I, 50.5; II, 50.7; III, 50.3; IV, 50.7; 
V, 49.9; VI, 49.9; VII, 49.9; VIII, 50.3. In Tests I to V, inclusive, 
commercial kerosene was used, inasmuch as the presence of sodium 
bicarbonate during distillation would seem to render purification 
unnecessary, but purified kerosene was used for the remainder. In 
Tests IV, V, VI, and VIII a very loose plug of well-teased glass 
wool was placed in the neck of the flask below the side tube to insure 
against the entrainment of soap. The extreme absolute error of any 
determination was thus ±0.4 per cent. The separatory funnel is 
graduated only to 0.2 cubic centimeter and was read only to the 
nearest 0.05 cubic centimeter. Inasmuch as the aqueous liquids 
do not always leave a uniform film on the glass when drained from 
the funnel, it appears possible for a reading to be in error by ±0.05 
cubic centimeter. If the two readings on a determination are both 
subject to errors of 0.05 cubic centimeter, the total error of reading 
may be 0.10 cubic centimeter, corresponding to 0.4 per cent of the 
sample. Thus the observed error of the determinations is practi- 
cally accounted for by the inherent error of the volumetric appa- 
ratus and the manner in which it is necessarily employed. A plug 
of glass wool appears superfluous, but it is probably best as a matter 
of principle to use only purified kerosene. 
As a result of the foregoing work the writer recommends the fol- 
lowing methods for the estimation of total phenols in commercial 
cresol and in saponified cresol solutions : 
APPARATUS 
Weiss tar-acid separatory funnel, type 2; 200 cubic centimeter 
hard-glass distilling flask ; vertical water-cooled condenser ; shield 2 
for flask; standardized 25 cubic centimeter pipette; standardized 25 
cubic centimeter cylinder. 
REAGENTS 
Sulphuric-acid solution, specific gravity 1.50-1.51 at 25° C; so- 
dium-hydroxide solution, substantially free from carbonate, specific 
gravity 1.10-1.11 at 25° C. ; sodium bicarbonate in powder, free 
from lumps; kerosene, recently purified by being washed twice with 
one-tenth of its volume of sodium-hydroxide solution, then with 
water until neutral, and freed from water by settling and filtration. 
PROCEDURE FOR CRESOL 
The pipette is filled above the mark with the sample at 25° C, 
its stem wiped dry, the contents brought to the mark and then deliv- 
2 The most convenient shield is made from hard glass, and may be extemporized by 
removing the bottom from a hard-glass beaker of appropriate size. In other respects the 
distillation apparatus prescribed by Weiss is admirable. 
