CONTROL OF SAPONIFIED SOLUTIONS 19 
possess an orange or red tint. From each of the two benzophenol 
tubes 20 cubic centimeters will be pipetted into 100 cubic centimeter 
volumetric flasks, treated with 5 cubic centimeters of dilute nitric 
acid, made to the mark, and mixed. The red flask contains the 
''phenol standard"; the yellow the "phenol blank." These phenol 
solutions are transferred to burettes. Ten cubic centimeters of each 
sample solution are pipetted into a Nessler tube. The orange con- 
stitutes the "-unknown" and the yellow the "sample blank," and 
each Nessler tube must be distinctly marked thus to avoid confusion 
later. Now, into the " sample blank " is run a measured amount of 
phenol standard from its burette, and the same volume of phenol 
blank is run into the " unknown "; both tubes are well agitated, aided 
by the insertion of rubber stoppers, if necessary, and the colors com- 
pared. When the tubes have been brought directly to a match each 
1 cubic centimeter of the phenol standard employed is equivalent to 
2 per cent of benzophenol if the original sample was 2.5 grams of a 
commercial cresol, or to 1 per cent if it was 5 grams of a saponified 
cresol solution. 
Precautions. — A pair of phenol tubes affords sufficient final solu- 
tions for assaying several unknowns, but all the latter must have ac- 
companied the phenol solutions throughout the entire process with 
identical reagents and treatment. If the end point has been inad- 
vertently overrun it is possible to work back to it ; but, since mistakes 
are easy to make in this procedure, it is better to repeat the compari- 
son on fresh portions from the original tubes. Too much delay in 
matching the tubes must be avoided once the titration has been 
started, else the excess of formaldehyde yet present in the blanks 
may have time to affect the intensity of the red after admixture. 
Millon solution is too dangerously poisonous and corrosive to justify 
the risk of using an ordinary pipette with suction by the unpro- 
tected mouth for its transference. 
It may be noted that the quantity of the phenols entering a single 
test, about 1.25 milligrams, is much less than the 4 milligrams em- 
ployed in the original paper. (2) It was then shown that the 
method gave distinctly low results, supposedly owing to the effect of 
formaldehyde. Later investigations have indicated the low results 
to be, in fact, primarily due to the adsorption of the red color by the 
yellow precipitates produced from the higher homologues of phenol. 
As the concentration of the total phenols in the sample is reduced, 
these precipitates become of less significance and the results for 
benzophenol approach the theoretical, as will be evident from ex- 
periment 8 correlative with experiments described in the paper 
referred to (;?). 
EXPERIMENT 8 
A solution of cresol was prepared by blending equal volumes of 
0=, m=, and p= cresol of "highest purity," and dissolving in 
water. Portions of this solution were mixed with an aqueous solu- 
tion of pure benzophenol in proportions to afford mixtures of known 
concentration in total phenols and in benzophenol, and these* mix- 
tures were assayed- The concentration of benzophenol in the stand- 
ard solution was in each assay equal to the concentration of the total 
phenols in the sample. The work and its results &re sufficiently ex- 
plained by the data contained in Table 4. 
