COXTKOL OF SAPONIFIED SOLUTIONS 17 
as soon as all water has passed off. The flame toward the end should 
be slightly luminous to mitigate local overheating of the flask. A 
250-cubic centimeter flask may be used in lieu of the 200-cubic centi- 
meter size. In dealing with opaque samples it will be necessary 
either to graduate a measuring cylinder to contain 25 cubic centi- 
meters at the top of the meniscus, using for the purpose a transparent 
sample of a saponified cresol solution, or to determine the specific 
gravity of the sample and then to weigh out the equivalent of 25 
cubic centimeters. Inasmuch as 1 cubic centimeter of water at 25° 
C. weighs 0.996 gram, the weight of 25 cubic centimeters of the 
sample will be 25 X specific gravity X 0.996 gram. 
The possible seriousness of one weakness in the method may be 
determined only by subsequent extended practical experience. The 
quantitative recovery of the total phenols becomes more doubtful as 
their boiling points rise and as they are mixed with increasing pro- 
portions of soap. However, a limit is set to the proportion of soap 
by the necessity for passing the chill test and the solution-rate test, 
and a limit is set upon the proportion of inordinately high-boiling 
jmenols by the necessity for passing the solubility-degree test. In 
view of these restrictions it is anticipated that no serious deficiency 
in the quantitative recovery of total phenols will be experienced in 
case of such samples as are, in other respects, fit for actual use. 
The results are obtained in volume percentages but these are the 
same within the inherent limits of error of the method as the weight 
percentages, for the specific gravity of a saponified cresol solution dif- 
fers but little from the specific gravity of the phenolic mixture used 
in its preparation. In experiment 6 the specific gravity of cresol 
H B was stated to be 1.0225, while that of saponified cresol solution 
H B was 1.0259. Cresol L B possessed a specific gravity of 1.034, 
while that of a saponified cresol solution prepared therefrom was 
found to be 1.030. 
With respect to the assay of commercial cresol, it is possible that 
if sodium bicarbonate is not added and if the kerosene is first dried, 
a sufficiently accurate estimation of tar bases plus water may be 
made by measuring the volume of distillate before and after it is 
shaken with sulphuric-acid solution. 
ESTIMATION OF BENZOPHENOL 
The writer (2) has previously described a method for the estima- 
tion of benzophenol in the presence of its homologues; accordingly, 
the present discussion will be limited to the essentials necessary for 
the routine execution of the method. 
REAGENTS 
Dilute nitric acid is prepared by blowing air through C. P. con- 
centrated nitric acid until colorless, then diluting 1 volume with 4 
volumes of water. Milloms reagent is prepared from 2 cubic centi- 
meters of mercury and 20 cubic centimeters of C. P. concentrated 
nitric acid in a 100 cubic centimeter Erlenmeyer flask placed under a 
hood and shaken as much as need be after the first violent reaction to 
effect subdivision of the mercury and maintain action. After about 
