112 The N.Z. Journal of Science and Technology. 
[April 
siliceous sand, 
but at certain points is 
concentrated by the action 
of water 
and wind to a 
rich deposit of the following average analysis :— 
Per 
Per 
Cent. 
Cent. 
Metallic iron 
. . 57 
Lime 
. . 2-16 
Phosphorus 
. . 0-25 
Mn0 2 
. . 0-64 
Sulphur. . 
. . 002 
MgO 
. . 2-8 
Titanium 
6 
Ti0 2 . . 
.. 9-31 
Silica 
. . 2-2 
P 2 0 5 .. 
. . 06 
Alumina 
. . 3-3 
v 2 0 5 .. 
. . 0-28 
Ferric oxide 
. . 44-36 
80 3 .. 
.. 0-05 
Ferrous oxide 
. . 33-91 
These rich deposits might form a continuous source of supply, the 
withdrawals being covered by fresh deposits, but this is doubtful; in any 
case, the sands may be said to be inexhaustible if New Zealand’s needs only 
are considered. 
Fig. 2.—-View of the smelting-works. 
The ironsand is very similar in analysis to mill scale, the gangue being 
practically all free. The ferrous-oxide content is unusually high, thus 
retarding reduction. Ferric oxides are usually largely reduced by CO, 
giving the exothermic reaction Fe 2 0 3 + 3CO = 2Fe + 3C0 2 with emission 
of heat. But ferrous oxides generally require a large proportion of fixed 
carbon to effect their reduction, giving the endothermic reaction FeO -f- C 
= Fe -j- CO with absorption of heat. Ores with high percentage of FeO 
therefore require large quantities of fuel and great length of time for 
reduction. The attempt was therefore made to reduce the ironsand largely 
by direct carbon, and the idea of making a “ ferro-coke ” by coking in a 
fireclay retort for eight hours equal parts of dried magnetically-concentrated 
sand and of coking-coal was evolved. This ferro-coke was charged along 
with limestone in a small experimental blast furnace, 3 ft. in diameter and 
8 ft. from changing-door to tuyeres, with a blast pressure of 6 oz. This 
process gave 10 tons per day of a very close-grained grey iron, con¬ 
taining 2-82 per cent, graphitic carbon, 075 per cent, combined carbon, 
