[STANSFIELD- TITANIFEROUS ORES OF IRON 39 
WISSLER] 
decided to simplify the investigation by keeping the alumina constant 
at ten per cent-and varying the silica, lime and titania. The method 
adopted was that of making small test cones of the various mixtures 
and determining the temperature at which each cone bent over when 
heated in a furnace. 
The materials from which the cones were made were as pure as 
could be obtained. The silica was prepared by crushing quartz until 
it would pass through a sieve of 100 meshes to the linear inch, and 
washing with hot acids until it was free from soluble matter. Alum- 
ina was made by adding ammonia to a solution of aluminium sulphate, 
washing the precipitate and igniting it in-a muffle at a bright red heat. 
Titanium oxide was obtained as the chemically pure reagent, and was 
tested, with negative results, for all the common impurities. Lime 
was prepared by heating chemically pure calcium oxalate in a muffle 
until it was free from volatile matter. All materials were tested for 
purity and were kept under such conditions that neither moisture nor 
carbon dioxide could be absorbed. 
The ingredients were weighed out, in the different proportions, 
so that the mixture in each case amounted to five grams; an amount 
which was sufficient tor two cones. The materials were ground 
together in an agate mortar, while dry, until thoroughly mixed, and 
then moistened with water toa stiff paste. It was found that mixtures 
containing 30% or more of lime could not be moulded directly into 
cones, as they would crack and fall to pieces after drying; owing no 
doubt to the absorption of carbon dioxide by the lime. To prevent 
this trouble the moistened mixtures were placed on glass plates and 
were allowed to dry in the air for several days. They were then re- 
ground, moistened and moulded into cones. 
The “cones” are three-sided pyramids, like the well known Seger 
cones, and were formed without difficulty in a mould made by cutting 
a V-shaped groove in a block of wood. The wooden mould was 
coated with melted paraffin and was lightly oiled with a cloth before 
each cone was made. The paste was pressed into the groove, smoothed 
off with a spatula, and pushed out by using the bottom of a Seger cone. 
The cones were then allowed to dry in the air or in an oven. 
The melting point tests were made in an alundum muffle heated 
by a Meker gas burner; but for temperatures above 1,500°C an 
electrically heated carborundum muffle was employed. 
The temperature measurements were made with a platinum, 
platinum-rhodium thermocouple and also with Seger cones. For 
experiments of this kind the Seger cone indications are accepted as 
being more satisfactory than the actual temperature readings as shown 
by the thermo-couple, and the results reported in this paper are based ° 
