6 BULLETIN" 1308, U. S. DEPARTMENT OF AGRICULTURE 
sary, in an appropriate cell under a compound microscope, Occa- 
sional particles of solid matter are of no significance, but globules 
should be entirely absent. It must be remembered that this test is 
made with distilled water, so that even if very few and very minute 
globules appear it may be presumed that the product will suffer 
pronounced breaking when mixed with the harder waters employed 
in the field. 
THE ESTIMATION OF TOTAL FIXED ALKALI 
The method here recommended is based upon one in common use 
for determining the total alkali in hard soaps, whereby a hot aqueous 
solution of the sample is treated with standard mineral acid in ex- 
cess, the solidified cake of fatty acids removed after cooling, and 
the excess acid titrated back with standard alkali and an appropriate 
indicator, such as methyl orange. Two additional steps are needed 
to make the method applicable to saponified cresol solutions — tar 
bases must be first expelled by evaporation, inasmuch as they are 
alkaline toward methyl orange, and stearic acid must be added in 
order to stiffen the separated fatty acids and cresol sufficiently to 
form a cake. 
Approximately 10 grams of the sample are weighed by difference 
into a 200 cubic centimeter beaker and reduced nearly one-half, or 
until no odor of tar bases is perceptible, by evaporation on a steam 
bath or on a hot plate, provided the temperature is maintained low 
enough to avoid decomposition which would produce a discolored 
solution for final titration. The residue is dissolved on the steam 
bath in about 100 cubic centimeters of water, then half -normal min- 
eral acid is run in from a burette until a drop of the well-stirred 
liquid produces a deep blue spot upon Congo red test paper. About 
5 grams of stearic acid is added and the beaker is heated upon the 
steam bath until the stearic acid is entirely melted, when the whole 
is vigorously stirred, left about 15 minutes for the layers to separate, 
and then cooled until the fatty acids form a firm cake. After filtra- 
tion, during which the filtrate is repassed if necessary to obtain per- 
fect clarity, and washing with cold water, the filtrate is titrated with 
half -normal alkali and with methyl orange as indicator. From the 
net consumption of standard acid is calculated the percentage of 
total alkali in the sample, expressed as sodium hydroxide. The 
stearic acid used must be neutral as respects methyl orange, but 
need not otherwise be of highest purity. 
THE ESTIMATION OF EXCESS FIXED ALKALI 
The excess alkali in ordinary soaps is commonly determined (7) 
through the titration of an alcoholic solution with phenolphthalein as 
indicator. That method can not be applied directly to saponified 
cresol solutions, because cresols in alcoholic solution react decidedly 
acid toward phenolphthalein. But if Poirrier blue C 4 B (3) be 
employed as indicator and the end point be observed in a special 
manner, the titration affords sufficiently accurate results. When soap 
only is present the color of the indicator undergoes a sharp change 
from cherry to blue, Avith sometimes a violet intermediate stage, at 
the same point at which phenolphthalein is changed from rose to 
