CONTROL OF SAPONIFIED SOLUTIONS 5 
(c) It shall contain soap, exclusive of excess alkali, equivalent to not 
less than per cent of sodium hydroxide, and the excess alkali 
shall be equivalent to not more than per cent of sodium hydroxide. 
(d) It shall contain not less than per cent and not more than 
per cent total phenols, not more than per cent of benzo- 
phenol. and not more than per cent of phenols distilling above a 
temperature of 
It will now be in order to formulate methods of test whereby the 
analyst may .determine whether a given sample meets the speci- 
fications. 
PHYSICAL TESTS 
The probable performance of complex materials in service may 
often be judged better by a few well-chosen physical tests than by 
data from a chemical analysis. An ideal physical test would sub- 
ject the sample to an exact replica of service conditions ; but in prac- 
tice it is usually necessary to adhere strictly to a set of more or less 
arbitrary conditions which imitate possible service conditions only 
partially and imperfectly. 
THE CHILL TEST 
A test tube, approximately 180 by 20 millimeters, is half filled 
with the sample, stoppered, and immersed at least three-fourths in a 
bath held at the desired temperature • for the desired period. The 
contents of the tube immediately after withdrawal from the bath 
should flow when the tube is tilted and should show no separation of 
soap, either in mass or as a pronounced turbidity. 
THE SOLUTION-RATE TEST 
An ordinary 250 cubic centimeter glass-stoppered graduated cylin- 
der about 35 centimeters in height and holding about 300 cubic centi- 
meters to the neck, is filled nearly to the neck with distilled water at 
the desired temperature and a single ordinary " marble " measuring 
12 to 14 millimeters in diameter is dropped in. Ten cubic centi- 
meters of the sample, free from froth, is rapidly poured from a 
graduate into the cylinder without touching its walls. The cylinder 
is quickly filled to overflowing with more distilled water at the same 
temperature and the stopper is inserted without entrapping any 
air; then the cylinder is held vertically between the hands and in- 
verted sharply once every 2.5 seconds — that is, brought through a 
complete cycle of positions in each 5 seconds. At the expiration of 
a fixed time counting from the first contact, the result should be a 
practically clear solution in which no jelly particles or globules of 
the undissolved sample are present. The test is valid only when the 
first inversion is made before any of the sample has fallen to the bot- 
tom of the cylinder and formed a layer thereon. 
THE SOLUBILITY-DEGREE TEST 
If the solution obtained in the solution-rate test is perfectly trans- 
parent and clear, the degree of solubility is satisfactory. Other- 
wise the solution is poured into a capacious flask, swirled a short 
time to insure solution of jelly particles, and then examined for 
globules, first with a hand magnifying glass, and finally, if neces- 
