38 THE ROYAL SOCIETY OF CANADA 
electrodes was about 450 amperes. The smelting began satisfactorily, 
but became slower owing to the accumulation of infusible materials 
in the furnace. The slag was barely fusible and there was also an 
excess of charcoal and of lime. The products were as follows: 
Slag analysis 
[SR V1 yee Be RER PRE ae Se Sts lbs. BIO r SEL nae ayers ae 24-51% 
SON locus been Meta ee 28 tbs. PO SIO shee Fee, be eae 
Limeand oreunfused: °.12+5 Ibs. “AWO3.... 4. eat ee Oe 
Gharcoale Mf Sn a 2 bs.) MCAaQ: foe, wen Fee NE À 
ME) mene te ar 3-12% 
FeO:-i ins ORNE 
99-12% 
The iron was white; but some of it, which had penetrated into 
the carbon lining of the furnace, was found, on fracture, to present the 
peculiar copper colour characteristic of the cyano-nitride of titanium. 
An analysis of this product was found to present certain difficulties, 
and has not been completed. 
The electrical energy consumed during the test amounted to 104 
kilowatt hours, apart from generator and transformer losses, which 
is equal to 3-4 K. W. hours per pound, or 1-17 h. p. years per long ton 
of iron. 
A rough balance sheet shows that the titania, alumina and silica 
in the charge are satisfactorily accounted for in the product obtained; 
the lime in the slag is, however, far less than in the limestone and ore 
charged, so that it is clear that there was a large excess of lime which 
remained in the unfused residue. There is also a loss of about one- 
third of the iron, which may be partly by volatilization and partly 
by filtration into the furnace bottom. The excess of charcoal left in 
the furnace shows that the bare amount calculated (21-4 lbs.) would 
have been ample without any allowance for mechanical losses. 
The unsatisfactory results obtained in this test threw doubt on 
the principles laid down by Rossi and it was decided to investigate the 
fusibility of titaniferous slags. 
Fusibility Tests 
These were made to ascertain the fusibility of mixtures of silica, 
alumina, lime and titania in varying proportions. Alumina in slags 
does not vary much and is frequently about ten per cent so it was 
