ASSAY OF ARSENICAL DIPPING FLUIDS. 13 
one-hundredth of 1 per cent of arsenious oxid in the bath under test. 
Therefore its strength should be originally fixed by standardizing it 
with the field apparatus against an average sample of used bath from 
the field in which the percentage (adjusted if necessary by the addi- 
tion of a little concentrated solution of arsenious oxid in sodium car- 
bonate) of actual arsenious oxid is close to 0.20 per cent and is accu- 
rately known through laboratory analysis. The true strength of this 
empirically standardized iodin solution should then be ascertained by 
titration against a strictly tenth-normal or twentieth-normal solution 
of arsenious oxid, and the result obtained may thereafter serve as the 
basis for the preparation of subsequent lots of test fluid. Obviously 
the true strength of the iodin solution will be influenced to some extent 
by the method of graduation of the cylinder, whether graduated "to 
contain" or "to deliver," and also by the depth of the meniscus, 
w T hich in turn is influenced by the diameter of the cylinder. Hence, 
all field outfits under the supervision of a single laboratory should be 
fitted with cylinders of uniform model. Reserve supplies of test fluid 
should be kept in small, well-filled, tightly closed glass-stoppered 
bottles, and in a cool, dark place. 
In addition to the test fluid, starch and sodium bicarbonate, or 
some equivalent substance, are of course necessary. In fact, the 
practical preparation of a satisfactory form of starch has been 
the greatest difficulty attached to the whole process, though at the 
same time the key to its success. 
It has been known for many years that by the use of alcohol starch 
may be precipitated in water-soluble form, also that high-percentage, 
yet mobile, starch solutions may be obtained through proper treat- 
ment with hydrochloric acid, but the working out of a practical 
process for the preparation in quantity of a dry starch readily soluble 
in cold water and appropriate for use as an indicator appears not to 
be recorded. 
Into a 5-liter round flask with a long neck is brought 400 grams potato 
starch, 2,300 c. c. distilled water, and, lastly, 80 c. c. of normal hydro- 
chloric acid. The flask is w T ell shaken to thoroughly wet and distrib- 
ute the starch and is floated in a kettle of water previously brought 
to vigorous boiling. The neck of the flask conveniently rests on the 
side of the kettle at an angle of about 45°, and as soon as the 
flask is brought into the bath it is gently but continuously rotated 
about its longitudinal axis. As the flask becomes hot the starch 
forms an evenly distributed, uniform jelly, which in about 7 minutes 
from the time of starting begins to liquefy and to fall away from the 
wall of the flask. When this stage is reached the mouth of the flask 
is loosely closed with an inverted beaker and the flask left in the boil- 
ing bath with an occasional rotation until the liquid becomes mobile 
