118 ' ANNUAL REPORT 
with hot water. Now wash the hydroxides from the filter into a clean 
beaker and dissolve in dilute sulphuric acid. In the presence of titanium, 
reduce the iron with hydrogen sulphide, which must be expelled by 
boiling and by passing carbon dioxide through the liquid. Now 
titrate the solution of ferrous sulphate with potassium perman- 
ganate as usual. The use of hydrogen sulphide is especially necessary 
‘in the analysis of fire clays, where titanium oxide is never lacking and 
the iron content is very low, while for brick clays it is of little importance 
whether this is done or not. The bulky colloidal precipitate of aluminum 
and ferric hydroxide can only be washed sufficiently by decantation and 
water. This is provided for by arranging a large can, made of tin-lined 
copper, holding from 5 to Io liters of water, at a considerable height above 
the laboratory table. This is kept gently boiling all the time and is sup- 
plied to the beakers or funnels by means of a long rubber tube provided 
at the end with a glass tube drawn out to a point and a pinch cock. This 
should be long enough to be swung from one end of the table to the 
other so that it may also be used for washing precipitates in place of the 
wash bottle. The tube can be counterbalanced by means of a weight run- 
ning over a pulley. 
The liquid to be decanted is best syphoned into an evaporating dish 
and kept boiling so that the volume may be reduced as quickly as possible 
to the desired volume. It is necessary to filter the concentrated washings 
in order to recover any small amounts of alumina and iron carried over. 
In this as well as other industrial chemical work it is of extreme 
importance to watch closely the quality of the reagents used. It is best 
to frequently run blank determinations on each process so as to ascertain 
the exact amount of the impurities being brought in by the reagents used. 
PROXIMATE OR MINERAL ANALYSIS OF CLAY 
Repeated attention has been called to the mineral composition of 
clay, to its principal mineral constituents, clay substatsce, feldspar and 
quartz. The actual determination of these three constituents, according 
to Seger and Bischof, is carried out as follows: 
Take one gram of the clay, transfer it to a casserole and add 50 cc. of 
sulphuric acid. Heat till white fumes begin to be given off and continue the 
heating for 8 hours. The volume must be kept up by the occasional ad- 
dition of more sulphuric acid. After this heating cool the liquid and dilute 
with water to about 250 cc. Heat the diluted solution, and filter through 
a hardened filter and wash the precipitate, which has been transferred to 
the latter, with a hot solution of dilute hydrochloric acid (1:5) followed 
by washing with hot distilled water. Now carefully wash the precipitate 
back to the casserole and add 50 cc. of a solution containing 3 grams of 
sodium carbonate and % gram of caustic soda. Heat almost to boiling for 
30 minutes, and decant through the same hardened filter. Wash any pre- 
cipitate on filter back into casserole, into which pour 50 cc. of a solution 
