SUMED, CROMOGCISIE, 113 
bined water of the silicic acid. This simple method, of course, is applicable 
only to materials whose general composition is already known and can 
never be used as the sole method of comparing pozzuolanes. 
ANALYSIS OF FURNACE SLAGS. 
This method is quoted chiefly from‘ Notes on Metallurgical Analysis,” 
by Professor N. W. Lord, professor of metallurgy and mineralogy at the 
Ohio State University and consulting chemist of the Ohio Geological 
Survey. 
Weigh out one gram of the sample in a casserole, add 30 cc. of 
water and stir the slag to prevent caking. Now add 20 cc. of hydro- 
chloric acid and heat. Everything should dissolve except a few flakes 
of silica and possibly a little carbon and sulphur. Evaporate to dry- 
ness and heat in the oven to a temperature between 110 and 120° C. 
till all of the acid is expelled. Take up residue in water and hydro- 
chloric acid, filter off and transfer filtrate back to casserole for a second 
evaporation and heating. Take up in acid again as before, filter and wash. 
Ignite and weigh as silica. This residue is liable to contain traces of 
iron, titanium and spinel. It may be tested with hydrofluoric acid and 
any fixed residue deducted. 
The alumina is now determined in the filtrate from the silica, to 
which a few drops of nitric acid have been added. In the absence of 
much manganese this can be done by precipitation with ammonium 
hydroxide, taking care to have plenty of ammonium chloride present 
and to redissolve the first precipitate, which is likely to contain a little 
lime. The precipitate is washed free from chlorine and then trans- 
ferred to the filter. It may contain iron, phosphoric and titanic acid in 
the slag, but as. these are rarely present in noticeable quantities they may 
be neglected. If the slag contains much manganese the alumina must be 
separated from it by a basic acetate precipitation. ‘The precipitate is then 
redissolved in hydrochloric acid and the alumina now free from manganese 
is precipitated by ammonium hydroxide. This, however, is rarely neces- 
sary. 
The presence of much manganese also influences the determination 
of calcium. If, hence, more than a few tenths per cent. of manganese is 
presentadd2or 3 cc.of ammonium sulphide to the filtrate from the alumina. 
Heat to boiling and boil for three minutes. Cool the liquid and transfer 
to a 500 cc. graduated flask; 250 cc. of this volume can now be measured 
out by filtering through a dry filter into 250 cc. graduated flask. In 
this solution, representing half of the sample, the calcium and magnesium 
can be determined gravimetrically or volumetrically. 
If ammonium sulphide was used, add HCl to the filtered solution till 
neutral and then about 5 cc. in excess. Boil till the hydrogen sulphide 
is expelled, then add 0.5 gram of potassium chlorate and heat till the 
separated sulphur is dissolved. Now add ammonium hydroxide and pro- 
8—S. G. Bull. 3. 
\ 
