ON THE CRYSTALLINE SLAGS. 367 
Berzelius, 7#™¢ année, p. 127, and Graham’s Chemistry), 2°62 of the salt 
would contain 1°28 of alumina. 
6. Sulphate of lime 0°23. 
7. Phosphate of magnesia 0°19. 
8. Sulphur. This determination was made by Mr. Forbes by fusion with 
nitrate of potass and the carbonate of potass-soda mixture. 20°14 grains 
gave of sulphate.of baryta 0°87. The odour of hydrosulphuric acid could 
not be'detected by the action of hydrochloric acid upon the powder of the 
slag; but on suspending for some time a slip of moistened acetate of lead 
paper in a test tube containing some of the powder of the slag and hydro- 
chloric acid, it became slightly brown. Now, as the slag contains sesqui- 
oxide of iron, the hydrosulphuric acid liberated by the action of hydrochloric 
acid upon any sulphuret which may be present, would evidently be imme- 
diately decomposed, with the separation of free sulphur, by the sesqui- 
chloride of iron which would be formed at the same time; and, accordingly, 
free sulphur was always distinctly recognised by its odour in drying the silica 
obtained from slags of this group. It may perhaps appear somewhat remark- 
able that sulphuret of iron and sesquioxide should exist together in the same 
slag; the fact however is certain. It may be that the sulphuret is irregu- 
larly diffused, and is, as it were, entangled in the mass. No trace of sul- 
phuric acid could be detected by the addition of baryta water to the solution 
of the slag in hydrochloric acid, even after standing 24 hours. 
9. Sesquioxide of iron. About 10 grains were digested in dilute hydro- 
chloric acid in a well-stopped flask over the water-bath. The necessary 
precaution of previously filling the flask with carbonic acid was carefully 
observed. The clear supernatant liquor was decanted rapidly into a stoppered 
phial containing excess of the solution of chloride of gold and sodium. The 
phial was well-closed and left for several days. The metallic gold weighed 
3°93, which by the tables corresponds to 4°16 FeO. The filtrate was deprived 
of the excess of gold by digestion with oxalic acid. The iron was precipi- 
tated by hydrosulphate of ammonia; redissolved in hydrochloric acid; per- 
oxidized by chloride of potass, and precipitated by ammonia ; the precipitate 
was boiled with potass, redissolved in hydrochloric acid, and precipitated by 
succinate of soda with the usual precaution. The sesquioxide of iron weighed 
5°76, but from this must be deducted 4°16 FeO, estimated as Fe? O3=4:°63 ; 
5°76 —4°63=1°13 iron existing as Fe?O%. The ratio of FeO to Fe? O3 is 
4°16: 1:13. The total amount of iron estimated as Fe® O8 is 72-42 per cent. 
5°76 : 1°13 :: 72°42: 14°20. But 0°60 of S, as sulphuret, was present, which, 
by tables, gives 2:91 Fe? O%. The total amount of Fe?O* per cent. is 
1420+ 2:91 =17°11 . 72°42—17:11=55'31 Fe? O3=49°73 FeO, from which 
must be subtracted a quantity of Fe proportionate to 0°60 S. This quantity 
is 1:01—1-01 +0°60=1:61=quantity of FeS. 1-01 corresponds to 1°30 FeO. 
49°73 —1-30=48'43=total FeO per cent. 
Fe as FeO 48°43 
Fe as Fe? O3 17-11 
Fe as FeS 1°61 
10. Phosphoric acid. Weight of slag 20°10. Proceeded to obtain asolu- 
tion in hydrochloric acid free from silica in the usual way. The tartaric acid 
_ process was adopted. The whole was allowed to stand several days. The 
ammoniaco-magnesian phosphate was washed with cold ammonia water. 
Phosphate of magnesia 0°42. Colour pale brown from a trace of manga- 
hese. Every trace of carbon was burnt out. 
