Donovan. — TaranaJci Ironsand. 507 



Summary. 



The results of the investigation may be summarized as follows : — 



(1.) The percentage of iron in the magnetically separated concentrates 

 lies between 50 and 60 per cent., and under good working-conditions would 

 approach the latter figure. 



(2.) Vanadium varies with the iron-content — from 0-08 per cent, in the 

 poorer samples to 0-25 and 0-34 in the richer ores. The greater portion 

 of it remains with the iron when the sand is magnetically separated. Its 

 presence in such small quantities confers no advantage, as, on smelting, it 

 would pass into the slag, and not into the iron. But if means could be 

 devised to completely recover all the vanadium, and convert it into ferro- 

 alloy, 1 ton of sand containing 0-2 per cent, of vanadium would at the 

 present prices produce alloy worth approximately £2. 



(3.) The titanic oxide varies in the original sands from 6-2 to 10-6 per 

 cent., and remains with the strongly magnetic portion, being slightly 

 increased relatively to the iron. This confirms the results obtained by 

 W. Skey (33rd Annual Eeport of the Colonial Laboratory, N.Z., p. 17). 



(4.) The phosphorus in the untreated sands varies from 0-16 to 0-28 per 

 cent. It is not appreciably decreased, relatively to the iron, by magnetic 

 separation. Fine grinding prior to separation did not give any better 

 result. The phosphorus still remained in such quantities in the concen- 

 trated ironsand as would necessitate the use of the basic process for the 

 manufacture of steel from the ore. 



The results from a commercial standpoint are disappointing. They 

 indicate that the phosphorus and titanium are intimately associated with 

 the iron in the sand, and cannot be readily eliminated, if at all, by magnetic 

 means. It is possible, however, that a method will be found which will 

 achieve this end. 



In conclusion, I would acknowledge the ready assistance of Messrs. 

 N. L. Wright and E. P. Wilson in much of the analytical work involved. 

 I would also thank Dr. J. S. Maclaurin, Dominion Analyst, for permission 

 to publish these results. 



