1918.] The N.Z. Journal of Science and Technology. 45 
plished, but nothing successful was done in the way of producing iron 
beyond at times a cake of fritted wrought iron in the crucible. It was 
realized that the indirect heating of the charge by coke through the walls 
of the crucible in any case could not be applicable on a commercial scale, 
but enough confidence was established in the practicability of powdered- 
coal firing to warrant further trials. 
The second furnace built is shown in fig. 2. A larger blower was 
obtained, and the feed-mechanism changed to a screw. The air was divided 
and controlled by stop-cocks, so that the supply to the combustion-chamber 
could be completely regulated. The by-pass allowed all the air not going 
into the combustion-chamber to go to the melting-hearth. A comparatively 
low temperature and highly reducing conditions were thereby produced 
in the combustion-chamber, and it was hoped thereafter to melt the reduced 
ore by the intense heat produced in the melting-hearth. All that happened, 
however, was the reoxidizing of any reduced sand, and the formation of 
copious baths of melted magnetite. We struggled with this furnace long 
Fig. 2.—Second experimental furnace. 
and hard, but it was a hopeless task; and I am convinced it is not possible 
in any other way than by the opposed-stream method of the bias furnace 
to produce with fuel-fired furnaces a flame which shall be hot enough to 
melt the finely divided particles of reduced sand without at the same time 
reoxidizing them. According to what I have been able to read, this was 
the difficulty experienced in all direct processes of making wrought iron 
which were tried extensively about 1880. None proved successful com¬ 
mercially, though it was shown that by extreme care in manipulation—as, 
for instance, at Onehunga—malleable-iron blooms could be made. In the 
Blair furnace the iron sponge was cooled down in the lower part of the 
furnace by means of a water jacket; then withdrawn and compressed 
cold, with tar, into briquettes, which were then charged into an open-hearth 
furnace for the production of steel. 
The great readiness of the reduced iron-ore to reoxidize is due to its 
physical and chemical condition. It is almost pure iron, is not protected 
by contained carbon and silicon, and is in a spongy state, the particles of 
ore retaining their - original shape, but being perforated with innumerable 
holes. Iron presents several phenomena analogous to the reduction of 
the ore. Just as the ore can part with its oxygen while in the solid 
